Page 1

24366_int_Mise en page 1 22/12/10 12:25 Page1

JOURNAL OF MARKETING TRENDS - EDITORIAL BOARD Editors Jean-Claude ANDREANI ESCP Europe Paris Umberto COLLESEI Università Ca’ Foscari Venezia British Editor-in-Chief Christopher HALLIBURTON ESCP Europe London British Editorial Board Ruth ASHFORD Manchester Metropolitan University Kim CASSIDY Lincoln University John DAWSON The University of Edinburgh John EGAN London South Bank University John ENSOR Edinburgh Napier University Chris FILL Portsmouth University Susan HART University of Glasgow Phil HARRIS Chester University Andrew LINDRIDGE The Open University Milton Keynes Dale LITTLER University of Manchester Phil MEGICKS University of Plymouth Adrian PALMER University of Wales Swansea Stuart ROPER University of Manchester Peter TURNBULL Birmingham University Jeryl WHITELOCK University of Bradford Chinese Editor-in-Chief Taihong LU Sun Yat-Sen University China Editorial Board Rong CHEN Tsinghua University Guoqun FU Peking University Jiaxun HE East China Normal University Guihua LI Nankai University Siqing PENG Peking University Tao WANG Wuhan University Xingyuan WANG Shandong University Yonggui WANG University of International Business and Economics French Editor-in-Chief Jean-Louis MOULINS Université Aix-Marseille II France Editorial Board Philippe AURIER Université de Montpellier Christophe BENAVENT Université Paris X Dominique BOURGEONRENAULT Université de Bourgogne Joël BREE Université de Caen/ Rouen Business School Jean-Jack CEGARRA Université Lyon III Jean-Louis CHANDON Université Aix-Marseille III Gérard CLIQUET Université Rennes I Bernard COVA Euromed Marseille Jean-Marc DECAUDIN Université Toulouse I Pierre DESMET Université de Paris Dauphine / ESSEC Paris Pierre-Louis DUBOIS Université Paris II Marc FILSER Université de Bourgogne Jean-Luc GIANNELLONI Université de Savoie

Denis GUIOT Université Paris Dauphine Patrick HETZEL Ministère de l’Enseignement Supérieur Joël JALLAIS Université Rennes I Alain JOLIBERT Université de Grenoble Richard LADWEIN Université Lille I Jean-François LEMOINE Université Paris I Christian PINSON INSEAD Fontainebleau Bernard PRAS Université Paris Dauphine / ESSEC Paris Gilles ROEHRICH Université de Grenoble Elyette ROUX Université Aix-Marseille III Francis SALERNO Université Lille I Jean-françois TRINQUECOSTE Université Bordeaux IV Pierre VALETTE-FLORENCE Université de Grenoble Eric VERNETTE Université Toulouse.I Monique ZOLLINGER Université de Tours German Editor-in-Chief Klaus-Peter WIEDMANN Universität Hannover Germany Editorial Board Daniel BAIER Brandenburgische Technische Universität Cottbus Doris BERGER IMC Fachhochschule Krems Manfred BRUHN Universität Basel Hartmut H. HOLZMUELLER Universität Dortmund Frank JACOB ESCP Europe Berlin Michael KLEINALTENKAMP Freie Universität Berlin Michael LINGENFELDER Philipps-Universität Marburg Helge LÖBLER Universität Leipzig Roland MATTMULLER European Business School Oestrich-Winkel Anton MEYER Universität München Hans MUHLBACHER Universität Innsbruck Günter SILBERER Universität Göttingen, Thorsten TEICHERT Universität Hamburg Thorsten TOMCZAK Universität St-Gallen Kai-Ingo VOIGT Universität Erlangen-Nürnberg Jan WIESEKE Universität Bochum Gerhard WUHRER Universität Linz Indian Editor-in-Chief Abraham KOSHY IIM Ahmedabad India Editorial Board Saroj Kumar DATTA IIT Kharagpur Krishna HAVALDAR ABA Karnataka Nagasimha KANAGAL IMM Bangalore Ashok KAPOOR MDI Gurgaon Anjur RAMAN IMT Ghaziabad Ravi SHANKER IIFT Bhawan New Delhi Rajendra Prasad SHARMA IIT Kanpur

Italian Editor-in-Chief Tiziano VESCOVI Università Ca' Foscari Venezia Italy Editorial Board Claudio BACCARANI Università di Verona Giuseppe BERTOLI Università di Brescia Aldo BURRESI Università di Firenze Tiziano BURSI Università di Modena e Reggio Emilia Sergio CHERUBINI Università di Parma Guido CRISTINI Università di Padova Daniele DALLI Università di Pisa Mariangela FRANCH Università di Trento Gianpiero LUGLI Università di Parma Alberto MATTIACCI Università di Roma La Sapienza Alberto PASTORE Università di Roma La Sapienza Tonino PENCARELLI Università di Urbino Carlo Bo Luciano PILOTTI Università di Padova Sergio SCIARELLI Università di Napoli Federico II Paolo STAMPACCHIA Università di Napoli Federico II Enrico VALDANI Università Luigi Bocconi Milano Riccardo VARALDO Scuola Normale di Pisa Giuseppe VOLPATO Università Ca' Foscari Venezia Portuguese Editor-in-Chief Paulo de LENCASTRE Universidade Católica Portuguesa Portuguese Editorial Board Joana CÉSAR-MACHADO Universidade Católica Portuguesa Ana CÔRTE-REAL Universidade Católica Portuguesa Pedro DIONÍSIO ISCTE Lisboa Alexandre LAS CASAS Pontifícia Universidade Católica de São Paulo José PALANDI Pontifícia Universidade Católica de São Paulo Clotilde PEREZ Universidade de São Paulo Cláudia SIMÕES Universidade do Minho Ana Maria SOARES Universidade do Minho Geraldo TOLEDO Universidade de São Paulo Spanish Editor-in-Chief José María MUGICA Universidad Pública de Navarra Spain Editorial Board Manuel ALFARO ESADE Barcelona José Manuel BARREIRO FERNANDEZ Universidad de Santiago de Compostela Asunción BEERLI PALACIO Universidad de Las Palmas de Gran Canaria Carmen BERNE MANERO Universidad de Zaragoza Martina G. GALARZA Universidad Catolica de Valencia Irene GARCIA URETA Universitad del Pais Vasco Irene GIL SAURA Universidad de Valencia Ana-Isabel JIMENEZ-ZARCO Universitat Oberta de Catalunya

Nora LADO COUSTÉ Universidad Carlos III de Madrid Antonio LADRON DE GUEVARA Universidad Barcelona Teodoro LUQUE MARTINEZ Universidad de Granada Enrique MARTIN ARMARIO Universidad de Sevilla Josefa D MARTÍN SANTANA Universidad de Las Palmas de Gran Canaria Felix MARTIN VELICIA Universidad de Sevilla Maria Pilar MARTINEZ-RUIZ Universidad de Castilla-La-Mancha Francisco José MAS RUIZ Universidad de Alicante José Luis MÉNDEZ GARCÍA DE PAREDES Universidad Autonoma de Madrid Miguel Angel MOLINER TENA Universitat Jaume I Madrid Elena OLABARRI FERNANDEZ Universidad del País Vasco Javier OUBIÑA BARBOLLA Universidad Autonoma de Madrid José Luis PLACER GALAN Universidad de León Yolanda POLO REDONDO Universidad de Zaragoza Ignacio A RODRÍGUEZ DEL BOSQUE Universidad de Cantabria

International Editors Jean-Claude ANDREANI ESCP Europe Paris Umberto COLLESEI Università Ca’ Foscari Venezia International Editorial Board Sanem ALKIBAY Gazi University Michelle BERGADAA University of Geneva Jean-Charles CHEBAT HEC Montréal Yuxin CHEN New York University François H. COURVOISIER Haute École de Gestion Neuchâtel Patrick DE PELSMACKER University of Antwerp Torben HANSEN Copenhagen Business School Agnes HOFMEISTER Corvinus University of Budapest J. Jeffrey INMAN University of Pittsburgh Mustafa KOKSAL Kocaeli University Kemal KURTULUS Istanbul University Panagiotis KYRIAZOPOULOS Institute of Piraeus Athens Eric LI Eastern Mediterranean University Cyprus Raymond McLEOD Southwestern University Goergetown Patrick MURPHY University of Notre Dame Laura OSWALD University of Illinois Bogna PILARCZYK Poznan University of Economics Irene ROZEN Hub University College of Brussels Donald SEXTON Columbia University New York Henrikki TIKKANEN Helsinki School of Economics M. Mithat UNER Gazi University J.Miguel VILLAS BOAS University of California Demetris VRONTIS Intercollege Cyprus Patrick VYNCKE Ghent University

Journal of Marketing Trends - Volume I (February 2011) 1


24366_int_Mise en page 1 22/12/10 12:25 Page2

JOURNAL OF MARKETING TRENDS - REVIEWING COMMITTEE Marie-Hélène ABBO ESC Pau Ahlem ABIDI-BARTHE European Business School Paris Jadwiga ADAMCZYK Cracow University of Economics Jamal AHMAD Cardiff University Gaetano AIELLO Università di Firenze Thérèse ALBERTINI Université de Corse Manuel ALFARO ESADE Barcelona Sanem ALKIBAY Gazi University Manuel Angel ALONSO COTO IE Business School Madrid Laure AMBROISE Université de Villeurbanne Jean-Claude ANDREANI ESCP Europe Paris Vanessa APAOLAZA IBANEZ Universidad del País Vasco Cécile ARNOULD ESC Bretagne Ruth ASHFORD Manchester Metropolitan University Philippe AURIER Université de Montpellier Maria AVELLO Universidad Complutense de Madrid Claudio BACCARANI Università di Verona Daniel BAIER Brandenburgische Technische Universität Cottbus Christine BALAGUE Université Lille I José Manuel BARREIRO FERNANDEZ Universidad de Santiago de Compostela Guy BASSET Université Rennes I Juan Manuel BATISTA ESADE Barcelona Asunción BEERLI PALACIO Universidad de Las Palmas de Gran Canaria Christophe BENAVENT Université Paris X Michelle BERGADAA University of Geneva Doris BERGER IMC Fachhochschule Krems Carmen BERNE MANERO Universidad de Zaragoza Giuseppe BERTOLI Università di Brescia Madeleine BESSON TELECOM & Management SudParis Enrique BIGNE ALCANIZ Universidad de Valencia Zeynep BILGIN Marmara University Andreu BLESA Universitat Jaume I Madrid Kristin BLONDE University of Antwerp Lionel BOBOT NEGOCIA Paris Philippe BOISTEL IAE de Rouen

Joaquim BORGES GOUVEIA Universidade de Aveiro Denis BORIES Université Toulouse II Jean-François BOSS RFM Paris Patrick BOUCHET Université de Bourgogne Othman BOUJENA ESC Rouen Laurent BOURDEAU Université Laval, Quebec Dominique BOURGEONRENAULT Université de Bourgogne Neji BOUSLAMA Université de Tunis Cécile BOZZO Université Paris Dauphine Joël BREE Université de Caen Gregory BRESSOLLES BEM Bordeaux Management School Manfred BRUHN Universität Basel Michael BRUSCH Brandenburgische Technische Universität Cottbus Isabel BUIL CARRASCO Universidad de Zaragoza Aldo BURRESI Università di Firenze Tiziano BURSI Università di Modena e Reggio Emilia Bruno BUSACA Università Luigi Bocconi Milano Silvia CACHO ELIZONDO EBS Paris Michel CALCIU Université Lille I Haydee CALDERÓN GARCÍA Universidad de Valencia Béatrice CANEL-DEPITRE Université Le Havre Francesco CASARIN Università Ca'Foscari Venezia Daniel CAUMONT ICN Business School Marie-Cécile CERVELLON International University of Monaco Amparo CERVERA TAULET Universidad de Valencia Jean-Louis CHANDON Université Aix-Marseille III Chandan CHATTERJEE MICA Ahmedabad Jean-Charles CHEBAT HEC Montréal Rong CHEN Tsinghua University Yuxin CHEN New York University Sergio CHERUBINI Università Tor Vergata Roma Rosa CHUN University of Manchester Gérard CLIQUET Université Rennes I François CODERRE Sherbrooke University Enrico COLLA NEGOCIA Paris Umberto COLLESEI Università Ca'Foscari Venezia

2 ISSN 1961-7798 - © 2009, International Marketing Trends Conference

Ana CORTE-REAL Universidade Católica Portuguesa Thomas COSSE Richemond University Patrice COTTET Université de Reims François COURVOISIER Haute École de Gestion Neuchâtel Bernard COVA Euromed Marseille Cécile CRISTAU Université de La Rochelle Jean-Philippe CROIZEAN Université Rennes I Daren DAHL Vancouver University Daniele DALLI Università di Pisa Leila DAMAK Université Bretagne Sud Mihir DASH ABA Bangalore Saroj Kumar DATTA IIT Kharagpur Patrick DE PELSMACKER University of Antwerp Mozhen DEBBABI ESC Manouba Tunis Jean-Marc DECAUDIN Université Toulouse I / ESC Toulouse Sihem DEKHILI Université de Strasbourg Philippe DESHAYES Ecole Centrale de Lille Pierre DESMET Université Paris Dauphine Barbara DI BERNARDO Università di Padova Tiéfing DIAWARA EDC Paris Delphine DION Université Paris I Rafaelle DONVITO Università di Firenze Françoise DOREY ESC Rouen Olivier DROULERS Université de Bretagne Sud Pierre-Louis DUBOIS Université Paris XII Christian DUSSART HEC Montréal Ida ERCSEY Széchenyi István University Pascale EZAN ESC Rouen André FADY Université Rennes I Claudia FANTAPIE ALTOBELLI Helmut-Schmidt-Universität Hamburg Anne-Laure FARJAUDON Université Paris Dauphine Jean-Marc FERRANDI ENITIAA Nantes Nathalie FLECK Université de Cergy-Pontoise Christophe FOURNIER Université Montpellier II Mariangela FRANCH Università di Trento Agnès FRANCOIS-LECOMPTE Université de Bretagne Sud

Azza FRIKHA ESC Manouba Tunis Jean FRISOU INSEEC Bordeaux Wolfgang FRITZ Technische Universität Braunschweig Guoqun FU Peking University Gorm GABRIELSEN Copenhagen Business School Jean-Philippe GALAN Université Toulouse I Martina G. GALARZA Universidad Catolica de Valencia Céline GALLEN Université de Nantes Julio GARCIA DEL JUNCO Universidad de Sevilla Irene GARCIA URETA Universidad del País Vasco Marie-Laure GAVARD-PERRET Université de Grenoble Diana GAVILAN Universidad Complutense de Madrid Jean-Luc GIANNELLONI Université de Savoie Irene GIL SAURA Universidad de Valencia Jean-Claude GILARDI Université Nancy II Georges GIRAUD ENITA de Clermont-Ferrand Magali GIRAUD Université Toulouse I Bruno GODEY ESC Rouen Miguel Angel GOMEZ BORJA Universidad de Castilla la Mancha Laurence GRAILLOT Université de Bourgogne Nathalie GUIBERT Université Paris II Guihua LI Nankai University Valérie GUILLARD Université Paris Dauphine Véronique GUILLOUX Université Paris XII Denis GUIOT Université Paris Dauphine Michael HAENLEIN ESCP Europe Paris Nibrass HAJTAIEB EL AOUD IHEC Sfax Christopher HALLIBURTON ESCP Europe London Torben HANSEN Copenhagen Business School Susan HART University of Glasgow Patrick HARTMANN Universidad del País Vasco Krishna HAVALDAR ABA Karnataka Jiaxun HE East China Normal University Oliver HEIL Joh. Gutenberg Universität Mainz Agnès HELME-GUIZON Université de Grenoble Nadine HENNIGS Universität Hannover


24366_int_Mise en page 1 22/12/10 12:25 Page3

JOURNAL OF MARKETING TRENDS - REVIEWING COMMITTEE Stéphanie HERAULT Université Paris I Rosa María HERNANDEZ MAESTRO Universidad de Salamanca Patrick HETZEL Ministère de l'Enseignement Supérieur Agnes HOFMEISTER Corvinus University of Budapest Hartmut H. HOLZMUELLER Universität Dortmund J. Jeffrey INMAN University of Pittsburgh Frank JACOB ESCP Europe Berlin Leila JAOUED Université Paris-Est Ana-Isabel JIMENEZ-ZARCO Universitat Oberta de Catalunya Anne JOYEAU Université Rennes I Anne JULIEN Reims Management School Nagashimha KANAGAL IIM Bangalore Andreas KAPLAN ESCP Europe Paris Ashok KAPOOR Institute of Gurgaon Slim KHALBOUS HEC Carthage Philip J. KITCHEN University of Hull Michael KLEINALTENKAMP Freie Universität Berlin Mustafa KOKSAL Kocaeli University Michaël KORCHIA BEM Bordeaux Management School Dominique KREZIAK Institut d'Études Politiques Strasbourg Tore KRISTENSEN Copenhagen Business School Kemal KURTULUS Istanbul University Joanne KUZMA Worcester University Panagiotis KYRIAZOPOULOS Institute of Piraeus Blandine LABBE-PINLON Audencia Nantes Christian LABORDE Université Toulouse II Nora LADO COUSTÉ Universidad Carlos III de Madrid Antonio LADRON DE GUEVARA Universidad Barcelona Richard LADWEIN Université Lille I Joëlle LAGIER ESC Rouen Christine LAMBEY-CHECCHIN Université de Clermont-Ferrand Caroline LANCELOT-MILTGEN Université d'Angers Claude LAVICKA Université Nancy II Laure LAVORATA Université Paris XII Lionel LE NIGNOL ESSCA Angers André LE ROUX Université de Poitiers

Anne-Marie LEBRUN Université de Bourgogne Laurence LECOEUVRE ESC Lille Jean-Marc LEHU Université Paris I Jean-François LEMOINE Université Paris I Eric LI Eastern Mediterranean University Cyprus Marie-Christine LICHTLE Université de Bourgogne Michael LINGENFELDER Philipps-Universität Marburg Helge LÖBLER University of Leipzig Carlotta LORENZO ROMERO Universidad de Albacete Fernando LOSADA PEREZ Universidad de Santiago de Compostela Leila LOUSSAIEF ISC Paris Taihong LU Sun Yat-Sen University Gianpiero LUGLI Università di Parma Teodoro LUQUE MARTINEZ Universidad de Granada Virginie MAILLE CERAM Nice Subhas MANAGULI MICA Ahmedabad Myriam MANZANO Université Paris V Gilles MAROUSEAU ENSIM Le Mans Jean-Louis MARTIN ENSAIT Roubaix Enrique MARTIN ARMARIO Universidad de Sevilla Josefa D MARTÍN SANTANA Universidad de Las Palmas de Gran Canaria Felix MARTIN VELICIA Universidad de Sevilla Eva MARTINEZ SALINAS Universidad de Zaragoza Maria Pilar MARTINEZ-RUIZ Universidad de Castilla-La-Mancha Mercedes MARTOS Universidad de Salamanca Francisco José MAS Universidad de Alicante Alberto MATTIACI Università di Roma La Sapienza Kevin MATTINSON Keele University of Newcastle Roland MATTMULLER European Business School Oestrich-Winkel François MAYAUX EM Lyon Raymond MENCARELLI Université de Bourgogne José Luis MÉNDEZ GARCÍA DE PAREDES Universidad Autonoma de Madrid Anton MEYER Universität München Miguel Angel MOLINER TENA Universitat Jaume I Madrid Jordi MONTANA ESADE Barcelona

Pierre MORA BEM Bordeaux Management School Christina MORENO TELECOM & Management SudParis Jean MOSCAROLA Université de Savoie Jean-Louis MOULINS Université Aix-Marseille II José María MUGICA Universidad Pública de Navarra Hans MUHLBACHER Universität Innsbruck Pablo A. MUNOZ GALLEGO Universidad de Salamanca Patrick MURPHY University of Notre Dame Antonio NAVARRO GARCIA Universidad de Sevilla Andrew NEWMAN University of Salford Greater Manchester Jean-François NOTEBAERT Université de Bourgogne Simon NYECK ESSEC Paris Renata OCZKOWSKA University of Cracow Hyunjoo OH University of Florida Elena OLABARRI FERNANDEZ Universidad del País Vasco Alain OLLIVIER ESCP Europe Paris Klaus OSTREICHER Worcester University Laura OSWALD University of Illinois Javier OUBIÑA BARBOLLA Universidad Autonoma de Madrid Gilles PACHE Université Aix-Marseille II Adrian PALMER University of Wales Swansea Gaëlle PANTIN SOHIER Université d'Angers Juliette PASSEBOIS Université Bordeaux IV Alberto PASTORE Università di Roma La Sapienza Pramod PATHAK Indian School of Mines Daniele PEDERZOLI ESC Rouen Tonino PENCARELLI Università di Urbino Carlo Bo Siqing PENG Peking University Jean-Marie PERETTI Université de Corse Rozenn PERRIGOT Université Rennes I / ESC Rennes Christine PETR LE HUEROU Université Rennes I Guenka PETROVA Medical University of Sofia Frédéric PICHON ESC La Rochelle Karine PICOT COUPEY Université Rennes II Bogna PILARCZYK Poznan University of Economics Luciano PILOTTI Università degli Studi di Milano

Christian PINSON INSEAD Fontainebleau Marie-Pierre PINTO Université de Limoges José Luis PLACER GALAN Universidad de León Fabrice POITEAUX Ecole Centrale de Lille Yolanda POLO REDONDO Universita de Zaragoza Bernard PRAS Université Paris Dauphine ESSEC Paris Isabelle PRIM ALLAZ Université Lyon II Maria PUELLES Universidad Complutense de Madrid Mathilde PULH Université de Bourgogne Anjur RAMAN IMT Ghaziabad Vithala RAO Cornell University Pedro REINARES Universidad Rey Juan Carlos Madrid María RIPOLLES Universitat Jaume I Madrid Philippe ROBERT-DEMONTROND Université Rennes I Ignacio A RODRÍGUEZ DEL BOSQUE Universidad de Cantabria Gilles ROEHRICH Université de Grenoble Javier RONDAN-CATALUNA Universidad de Sevilla Irene ROOZEN Hub University College of Brussels Dominique ROUX Université Paris XII Elyette ROUX Université Aix-Marseille III Kaouther SAIED BEN RACHED Université de Tunis Francis SALERNO Université Lille I Hector SAN MARTÍN GUTIERREZ Universidad de Cantabria Manuel J. SANCHEZ-FRANCO Universidad de Sevilla Matthias SANDER Universität of Konstanz Sergio SCIARELLI Università Federico II Napoli Christophe SEMPELS ESC Lille Gilles SERE DE LANAUZE IUniversité Montpellier II Donald SEXTON Columbia University New York Niousha SHAHIDI EDC Paris Ravi SHANKER IFT Bhawan New Delhi Narendra Kumar SHARMA ITT Kanpur Rajendra Prasad SHARMA IIT Kanpur Béatrice SIADOU MARTIN Université Montpellier II Günter SILBERER Universität Göttingen Rahul SINGH Birla Institute of Management Technology Lilia SMAOUI ISG Paris

Journal of Marketing Trends - Volume I (February 2011) 3


24366_int_Mise en page 1 22/12/10 12:25 Page4

JOURNAL OF MARKETING TRENDS - REVIEWING COMMITTEE Grazyna SMIGIELSKA University of Cracow Francesco SOLE PARELLADA Universidad Politécnica de Cataluña Lionel SOUCHET Université de Bourgogne Paolo STAMPACCHIA Università Napoli Federico II M.S. SUBHAS Karnatak University Dharwad Joan SUREDA Universidad de Barcelona Thorsten TEICHERT Universität Hamburg Elisabeth TISSIER-DESBORDES ESCP Europe Paris Torsten TOMCZAK Universität St-Gallen Maria TOROCSIK University of Pecs Vesselina TOSSAN EDC Paris Nadine TOURNOIS Université Nice Sophia-Antipolis

Jean-François TRINQUECOSTE Université Bordeaux IV Hemant TRIVEDI MICA Ahmedabad Peter TURNBULL Birmingham University Mithat ÜNER Gazi University Enrico VALDANI Università Luigi Bocconi Milano Pierre VALETTE-FLORENCE Université de Grenoble Josep-Francesco VALLS ESADE Barcelona Vina VANI MICA Ahmedabad Riccardo VARALDO Scuola Normale di Pisa Eric VERNETTE Université Toulouse I Tiziano VESCOVI Università Ca'Foscari VeneziaMarisa VILLANUEVA Universidad de Navarra, Pamplona

4 ISSN 1961-7798 - © 2009, International Marketing Trends Conference

J. Miguel VILLAS BOAS University of California Catherine VIOT Université Bordeaux IV Kai-Ingo VOIGT Universität Erlangen-Nürnberg Pierre VOLLE Université Paris Dauphine Giuseppe VOLPATO Università Ca'Foscari Venezia Demetris VRONTIS Intercollege Cyprus Patrick VYNCKE Ghent University Agnès WALSER LUCHESI Université de Strasbourg Tao WANG Wuhan University Xingyuan WANG Shandong University Yonggui WANG Peking University

Bart WEITZ University of Florida Klaus-Peter WIEDMANN Universität Hannover Jan WIESEKE Ruhr Universität Bochum Gerhard WUHRER Universität Hannover Maria Jesus YAGUE Universidad Autonoma de Madrid Hélène YILDIZ Université Nancy II Mark ZANDER Freiburger Institut für Musikermedizin Hongxia ZHANG Peking University John Z. ZHANG University of Pennsylvania Sondes ZOUAGHI Université de Cergy-Pontoise


24366_int_Mise en page 1 22/12/10 12:25 Page5

JOURNAL OF MARKETING TRENDS - EDITORIAL

The idea of developing a Marketing Trend Radar The more complex and dynamic the marketer’s world gets, the more we have to understand the underlying trends which mark the directions of future developments. This is true for mar-keting practitioners as well as for marketing scientists. The Journal of Marketing Trends (JMT) intends to take part in this debate by publishing articles which will highlight the new trends in our discipline while respecting the plurality and diversity of researches. Which fields of trends are relevant? 1. Regarding the boundaries of the marketing discipline, our objects of experience and knowledge, we have to highlight first of all the broadening and deepening of the marketing concept as two important mega trends which since late the 1960ies already had, but in future still will have an important impact on the development of the marketing thought. Broadening refers to the fact that besides companies in different industries more and more also other social entities (individuals, groups, organizations, institutions, cities, nations etc.) are using the ideas of marketing for designing their ex-change processes to market goods, services, ideas, places, persons, and organizations. Moreover, marketing’s focus was broadened to include a social entity’s transactions with all of its publics. Paying attention to all stakeholders and not only to the customers is already closely related to the deepening of marketing. Especially when interests of other stakeholders or, more in general, of the society in total are incorporated in designing the exchange processes with for instance customers. Deepening of marketing is here characterized by a more systemic view as well as by the perspective of social responsibility. To en-sure such an orientation in the case of companies, marketing was more and more de-veloped as a comprehensive concept of societal and strategic corporate leadership. 2. Along the line of making marketing science more scientific marketing thinking was profoundly dominated by the empiricist world view and the logical empiricist para-digm. In the meanwhile a lot of different scientific orientations (empiricism, construc-tivism, and criticism), paradigms (market, political economy, network business and the corporate social responsibility paradigm), metaphors (subjective world metaphor, liberating metaphor etc.), methodologies and methods of puzzle solving (quantitative vs. qualitative research, rsp. mixed research) are having an impact on marketing re-search. Such an epistemological differentiation might be treated as an important trend to pluralism in scientific marketing research. Also an expression of this trend is the elaboration of marketing theory along different perspectives: micro perspective = marketing of individuals or organizations, meso perspective = marketing from the per-spective of complex networks (distribution channels, supply chains, strategic alliances, industries etc.) or issues (realizing specific ecological and/or social goals), and macro perspective = analyzing marketing structures and processes as well as their impacts and consequences from a comprehensive societal perspective. 3. A third and of course very important field of marketing trends are those trends in the marketing environment which address a) general trends of societal change or more specific dynamics in economic, technological, socio-cultural, political-legal, and eco-logical conditions, b) trends in industry and/or market structures and processes, and c) trends regarding the conditions, requirements and expectations, thoughts and feelings, and especially behaviors of all stakeholders (customers, suppliers, retailers, competi-tors, consumer groups,

Journal of Marketing Trends - Volume I (February 2011) 5


24366_int_Mise en page 1 22/12/10 12:25 Page6

JOURNAL OF MARKETING TRENDS - EDITORIAL

media etc.). These trends are either discussed more in general or more specifically in view of relevant challenges (chances and risks) for marketers of all kind. The wide range of trends starts with mega trends like globalization, virtualization, growing international competition and ends with very concrete trends for in-stance regarding consumers’ information and buying behavior in the context of societal and market changes. 4. Of course, last but not least, those trends which describe the development of marketing management approaches of different marketers are of special interest. Partly they might only mirror already implemented marketing concepts as a reaction of marketing practice on environmental changes, partly the might also reflect normative concepts developed in marketing science to help marketers to meet relevant challenges in a better way. No matter if more descriptive or normative such marketing management trends range from comprehensive concepts of a marketing-oriented corporate leader-ship (corporate marketing) to new means of market research and the planning, imple-menting, and controlling of the 4P’s (product, price, place, and promotion). And in between trends in several management fields have to be taken into account. For in-stance, trends in the field of normative and strategic management like new approaches in defining corporate identity and culture, defining the business and business mission (new goal systems, concept of corporate social responsibility), designing positioning, targeting, branding as well as cooperation strategies (strategic alliances, efficient con-sumer response networks etc.). But also trends in the field of designing and imple-menting information systems, organizational processes and structures up to new re-ward systems which help to sensibilize the employees for copying with new chal-lenges and to secure the implementation of an economically very efficient as well as societal responsible corporate policy. In all fields of marketing trends the dialogue between marketing scholars as well as marketing practitioners of different institutions and countries should be advanced systematically. As part of the International Conference Marketing Trends, we have created research chairs and a research program called Marketing Trends Radar. The objective is to obtain a clear view rang-ing from a big picture to a more detailed understanding of the main trends both in the field of marketing science and of marketing management. This program is open to all researchers who wish to participate and are concerned with new trends and developments in marketing. In the context of the upcoming Marketing Trends Conference in Paris we will provide a detailed framework which structures the relevant fields of marketing trends and helps to file the differ-ent contributions already existing along the papers submitted to our former marketing trends conferences as well as those being submitted in future conferences. After having agreed upon a proper structuring of marketing trends we will start trying to registering and keywording existing papers. Regarding the input to the planned Marketing Trend Radar System the Jour-nal of Marketing Trends (JMT) will play an important role. Each issue of the Journal of Marketing Trends (JMT) will give voice to researchers as well as practitioners from all areas and continents who are developing knowledge on the major trends in marketing: new theoretical and practical marketing perspectives; new fields of original re-search in all aspects of marketing; cross-country and cross-industry research; research on the major changes in marketing paradigms; new marketing approaches.

Klaus-Peter Wiedmann, Jean-Claude Andreani, Umberto Collesei, Françoise Conchon, Chris Halliburton, Abraham Koshy, José Mugica, Paulo de Lencastre, Taihong Lu, Jean-Louis Moulins, Tiziano Vescovi

6 ISSN 1961-7798 - © 2009, International Marketing Trends Conference


24366_int_Mise en page 1 22/12/10 12:25 Page7

J O U R N A L O F M A R K E T I N G T R E N D S - P R O D U C T I N N O V AT I O N

Customer knowledge competence, consumer insight interpretive techniques and competitive advantage of the firm. When the emotional involvement of the consumer, in the process of purchase and consumption of the products, is very strong, and the product technology is quite homogenous among competing brands, the competence to integrate consumer knowledge within the firm and generate the consumer insights for marketing innovation becomes primary source of competitive advantage.

Keywords: Consumer insight, Customer knowledge, Value co-generation, Product innovation process, Qualitative marketing research, Consumer behaviour, Consumer interpretive techniques.

> Luigi Cantone University of Naples Federico II lcantone@unina.it

> Pierpaolo Testa University of Naples Federico II p.testa@unina.it

Journal of Marketing Trends - Volume I (February 2011) 7


24366_int_Mise en page 1 22/12/10 12:25 Page8

J O U R N A L O F M A R K E T I N G T R E N D S - P R O D U C T I N N O V AT I O N

1. Introduction and objectives The following paper arises from an ongoing research’s project focused on the interpretive consumer researches. The research’s aims are: 1. identifying, under a theoretical perspective, the new interpretive techniques of consumer insight exploration and emersion; 2. exploring, under an empirical perspective, how the consumers are able to acquire and exchange knowledge, experience and skills, and, on the other side, individually or collectively, innovating in order to satisfy current and new needs. This part of empirical research will be focused on specific categories of product and services, and will take place using a set of complementary qualitative techniques; 3. applying case study methodology (Yin, 2002; Flyvbjerg, 2006), in order to explore how leading firms, in the specific categories of products and services surveyed: a. perceive the role of consumer insight for the competitive advantage; b. understand the main techniques they use to explore it; c. translate the consumer insight in the marketing innovation process.

2. Theoretical background (and literature review) 2.1. The market knowledge and the customer knowledge competencies The importance for the firms of harnessing knowledge-based competence, which yields a competitive advantage, is well established in both the marketing (Day, 1994; Glazer, 1991) and strategy (Pralahad, Hamel, 1990) literature. Market knowledge and customer knowledge competence are two related concepts, yet separate (Campbell 2003, Li & Calantone,1998). Market knowledge competence refers to the processes that generate and integrate structured and organized market information in aggregate, which includes both customer and competitive information; customer knowledge competence, instead, refers to the processes that generate and integrate, inside the organization, market knowledge or information about specific customers. Campbell (2003) underlines the inimitability of customer knowledge competence in the market: “unlike customer knowledge, a customer knowledge competence is inimitable, because the processes of generating and integrating customer knowledge are embedded in organizational cognitive activities and are not observed readily from outside (Day, 1994; Pralahad, Hamel, 1990), and immobile, because these processes are created within the firm and cannot be purchased on the market (Day, 1994)”. In accordance with Covà (2003), the consumers have to “take the control” of the marketing variables traditionally predefined by the firms (distribution, product, service distribution, brand communication), using the consumers and/or helping them to develop their competencies, not learning on them but from them and their experiences.

8 ISSN 1961-7798 - © 2009, International Marketing Trends Conference

2.2. A deep and true understanding of the consumer: the customer insight Jeppesen (2001) argues, that consumers - over all times – are capable to acquire knowledge, experience and skills and individually innovating, in order to satisfy current wants and to materialize imagined possibilities. To involve the customer, in order to cooperate more efficiently and effectively in the knowledge co-creation process, will be one of the greater challenges for the management (Gibbert et al., 2002; Davenport et al., 2001; Lundvall, 1985; von Hippel, 1988). But, which could be a definition of customer insight? Sawhney (2004) defines it as: “act of intuitively seeing the inner nature of things”, and, more precisely, as “a fresh and not-yet-obvious understanding of customers that can become the basis for competitive advantage” of the firm. About customer insight, Zaltman (2003) says: "a lot goes on in our minds that we're not aware of. Most of what influences what we say and do occurs below the level of awareness. That's why we need new techniques: to get at hidden knowledge –insight to get what people don't know they know". The customer insight definition puts in evidence some distinctive characteristics (Sawhney 2004): a. it is a notyet-obvious discovery; b. it is a unique and fresh perspective; c. it rarely emerges from quantitative research; d. it is often rooted in an observed anomaly.

2.3. The methodology and the techniques of exploration of the consumer insight The generation of the firm theories is more often the result of the qualitative and conceptual researches, while the test of the theory more often is associated with the empirical validation of the hypotheses (Gummesson, 2005). Spanjaard and Freeman (2006) have put in evidence as the debate on the dominant paradigm (Kuhn, 1962) has had different evolutions over the time. Under the positivism perspective “quantitative methods take marketing from an art to a science, from conjecture to rigor” (Saunders, 1999). Like proposed by Hunt (1991), however, it is necessary to break down the wall of the dogmatism between positivism and interpretivism in the consumer researches, because aren’t so far as they seem and, in any case, he advocates a “critical pluralism”. In this paper we accept the research philosophy pointed out by Hunt (1991), and earlier by Siegel (1988) and Laudan (1977), who advocates “virtually every major period in the history of science is characterized both by the co-existence of numerous competing paradigms, with none exerting hegemony over the field, and by the persistent and continuous manner in which the foundational assumptions of every paradigm are debated within the scientific community”. Thanks to this opposite point of view, in the time has been asserted the conviction that greater they are the benefits when the two methodologies


24366_int_Mise en page 1 22/12/10 12:25 Page9

J O U R N A L O F M A R K E T I N G T R E N D S - P R O D U C T I N N O V AT I O N

are implemented together, with none prevailing on the other (Sawhney 2004, Gummesson, 2005). The quantitative researches would be positivist and deductive, finalized to confirm an insight through the test of the hypotheses and to validate or to reject a theory proposal or to acquire answers to a very defined research questions. They would be, moreover, methodologically rigorous and would tend to consider the truth like objective and universal. The qualitative market researches, instead, would be of interpretative and reactive nature; using an inductive understanding of the phenomena, they would be finalized to generate insights. Such researches are of qualitative and eclectic type, they consider the truth like subjective and social constructed. This “pluralism” suggests a balanced undertaking rather than a competitive one (Brannen 2003; Deshpande, 1983; Stafford and Stafford, 1993), improving the practice in the consumer researches. However, more often than not, the role of qualitative research, within a market research context, is limited to capturing, as an exploratory and inductive methodology, the key variables which are important for further scientific, quantitative and deductive scientific methods (Milliken, 2001), thus, underrating its potential (Hyde, 2000). In our opinion, the qualitative researches, mainly when of “projective” type and aimed to explore deep and hidden knowledge in the consumers, have an independent dignity and not always they can be integrated with quantitative researches of extensive type. In fact, they may be carried out two orders of considerations: 1. don’t lack, in managerial practice, and in the scientific research field (Spanjaard and Freeman, 2006), examples of adoption of qualitative methodologies with purpose of deductive type, that is confirmative of insight arisen in a precedent stage of the research, and pervasive of stages much more advanced in the cycle of new products and services development (confirmation of new product innovation, confirmation of a chosen about the variables of marketing mix); 2. when the insight exploration does regard high emotional involvement products and rooted and hidden aspects in the unconscious of the consumer, explored with methodology of “projective” inductive type, testing the arisen insight through quantitative marketing researches could not confirm the indications emerged in qualitative phase, not cause of the inexistence of the same, but of the unsuitability of the same quantitative methodologies to verify the validity and reliability of the same (Boddy 2005a, 2005b). In our opinion, the choice of the research methodology to adopt must be evaluated in relation: 1. to the level of existing knowledge on the object of study (if limited/or wide); 2. to the nature of the knowledge inquired (if superficial-aware or deep-hidden type); 3. in the perspective of the management, to the degree of innovation researched from the firm on the surveying object (incremental or radical). Generally, it can be said that the more the field of surveying is known, the knowledge inquired is of superficial or aware type, and the degree of innovation searched

by the firm is of incremental type (search of a fine tuning in the offer system or the variables of marketing mix), much more is useful, under these premises, apply a quantitative methodology to study (measure or explore) the phenomenon. When the surveying field is little known, the inquired knowledge is of deep or hidden type and the search for innovation (of product-service or of the business model) is of radical type, greater is the usefulness of qualitative market research and projective techniques. Moreover, in this last case, the insight deriving from the implementation of the qualitative techniques with exploratory purpose it is of strategic nature. In such cases it does not have sense to proceed to the confirmation of the same insight for quantitative way, as the questions through questionnaire with adoption of metric scale, cannot unhinge the psychological resistance of the consumer on aspects of which the same consumer often - like moreover indicated, from Zaltman (2008) - it does not have awareness.

3. Research methodology This paper presents the first findings of an ongoing research project on the consumer insight whose empirical surveying, according to the original design of research, involves twelve product/service categories and case studies of leading firms. For exploring the customer insights about the product-service categories surveyed, we have made use of three different and complementary qualitative market research techniques: focus groups (Morgan 1997 and Boddy 2005b); personal experience narrative by the telling story technique (Gabriel and Griffiths 2004, Chase, 2005, Pinnegar and Daynes 2007, Riessman 2008); sentence completion (Soley and Smith, 2008). The interviews were carried out in June and July 2009, at the Marketing Laboratory, Department of Management, Faculty of Economics, University of Naples Federico II. Using a pilot study approach, in this paper, we present the results of the empirical surveys relative just to four category field, on twelve surveyed: consumer goods-alimentary paste, electronics consumer goodsmobile phones, consumer goods-detersives, consumer goods-femme care. Therefore, the elaborations contained in this paper are relative to the outcomes of eight focus group (two focus for each product category analysed; a total of 64 consumers interviewed, 8 for each group involved), among the 24 totally realized on twelve categories, whose duration has been, approximately, of 60 minutes each. Have also been carried out six “personal experience narrative” interviews (telling story of the duration approximately of 30 minutes each) for every product category (a total of 24 consumers interviewed) and 112 written affirmations through the “sentence completion” for each products category (8 affirmations gathered from each of 14 respondents involved in the individual and focus group interviews) . The findings of the interviews based on the personal experience narrative (telling story) Journal of Marketing Trends - Volume I (February 2011) 9


24366_int_Mise en page 1 22/12/10 12:25 Page10

J O U R N A L O F M A R K E T I N G T R E N D S - P R O D U C T I N N O V AT I O N

have informed, in accordance with the principles of the grounded theory (Corbin and Strauss, 1990), the cognitive structure of the focus group. Respondents at the first phase of the research (six individual interviews) are different from those who took part to the focus groups (composed by 8 participants). The respondents have been responsible of purchase, female (for consumer goods pasta, detersive and femme care), with age greater of 35 years and, nearly all, in working conditions. For the product category mobile phone the respondents have been men and women with age greater of 24. To each respondent has been asked to tell, relatively to the product category inquired, its own personal history (and familiar for mobile phone, detersive and alimentary pasta) with reference to the principles that inspire: 1. the need (the alimentary style, the inter-personal communication, the home, the psychological well being) ; 2. the product purchase and/or consumption (or product use for category femme care); 3. the relationships with the distributive channels of such categories of product. During the eight focus groups (two for each product category) the chosen of the questions has been administrated considering the evidences arisen in the first stage of surveying and deepening the conceptual categories emerged during the personal experience narrative interviews. We present, also, in brief, the case studies of the following leading firms as: Barilla Spa, Fater Spa, Henkel Italia Spa, Nokia Italia Spa. The aim of this paper is to recognize the existence of insights concerning the products’ categories surveyed, using a complementary and integrated system of interpretive consumer research techniques. In other words, to verify the ability of such techniques to generate the insights not the reliability of these latter. These insights, indeed, could be tested, in a future phase of this ongoing research, using quantitative research methods.

4.1. The results of empirical survey on alimentary pasta on consumers With reference to the alimentary style of the interviewed ones, the analysis makes to emerge a young woman not good consumers of pasta like the man; also, because she gives excessive attention to the physical form, erroneously she attributes to the pasta a negative association regarding the diet. It is delineated, moreover, a search for greater pasta quality in particular for the familiar or convivial events that seem to assign to the pasta still a main role. Among other things the interviewed ones have nearly always marked the consideration of the smaller economic incidence of the pasta purchase cost on the total cost of the lunch also during the weekly days. Also, for those of greater quality, like the Gragnano ones, thirty/forty cents for person in a family of four components are not exceeded. It grows, in particular, the segment of quality pasta, and, from this point of view, Barilla is not

10 ISSN 1961-7798 - © 2009, International Marketing Trends Conference

considered such, being perceived under the quality of others brand of Italian industrial pasta (i.e. De Cecco, Pasta Garofalo). The interviewed ones have evidenced a different approach to the preparation of the pasta in the weekly days, and the festive ones; during the festivity and for the events, the search of elaborated dishes is more recurrent, like the pasta’s consumes of greater quality (“the Gragnano ones”). A third insight is inherent to a variable of the marketing mix, the packaging, that has a fundamental role for the customers. The packaging of Barilla, for example, does not like to the interviewed because the cardboard occupy an excessive space in the pantry, the smell of the cardboard is unpleasant, and it determines a bad conservation of the product (sometimes presence of micro organisms). If the confection breaths outside, it can be a limit when the conservation takes place in not optimal environment or atmosphere. In any case, the cardboard is of obstacle to the perception of the colour of the pasta and its “wrinkledness”, two features of the pasta’s quality. Some brands, in fact, use transparent or “naked” packaging, just to emphasize these characteristics of the pasta. The cardboard pack, likely, answers to requirements of optimal conservation of the pasta (the cardboard more than the plastic envelope breathes outside) and of positioning in the linear facing of shelf causes a different impact in terms of space and visibility of the confection. The last insight is not negligible for the aims of the advertising communication. The pasta, but more in general terms the cooking, assumes peculiar cultural traits in Southern Italy, ranging from the comparison with the values and the familiar equilibrium of the family of origin, to the enthusiasm of the cooks, not even housewives, who can achieve results of value in everyday cooking and freed themselves from the reference models.

4.2. The results of empirical survey on consumer electronic goods - mobile phone The empirical survey on the mobile phone purchasers has made to emerge some insights deserving of an eventual deepening: 1. the concept of the mobile phone as a gift is very strong in the perception of the interviewed. In fact, in particular for the adolescents, the first mobile phone represents always a gift; 2. the tendency not to switch off the mobile phone during the night, that often determines an its use like an instrument of lighting (a torch) for the nocturnal movements in house or in dark atmospheres; 3. the purchase of the mobile phone on Internet becomes more and more important above all in some segments of demand (the juvenile ones). Also, the social media marketing tools (forum, social network, blog, etc.) are useful in order to acquire technical information, opinions, references on products. It could be evaluated the opportunity to design branded instruments of social media marketing in order to avoid that the “voice” of the


24366_int_Mise en page 1 22/12/10 12:25 Page11

J O U R N A L O F M A R K E T I N G T R E N D S - P R O D U C T I N N O V AT I O N

consumers on Internet is left to the total initiative of the individuals; 4. the product is perceived like a “growing symbol” and that it contributes “to the construction of the own identity”. This insight could be interesting for a mobile phone communication strategy, in particular for the young consumers; 5. equally interesting, always to the aims of the communication strategy, is the metaphor emerged of mobile phone as “pendant to the neck”, by virtue of recalling the concept of precious accessory, but also of inseparability regarding the woman and fashion symbol.

4.3. The results of empirical survey on consumer goods-detersives The empirical survey on detersives has concurred to put in evidence some insights: 1. a strong attention to the environmental impact of the detersives that would push the users to apply them in thrifty way; 2. an important role, in many familiar groups, of the assistant to the domestic matters. This last one, being of fact the user of the products and also an expert, influences the choices and the perception of the products and, therefore, the adoption model in the middle and long period; 3. increasing attention to natural remedies as thought always effective: the alcohol, the vinegar, the ash, the ammoniac, etc. This is a field in which already there are the attempts of the leader brands, but still a lot is to be done; 4. the consumption of the unbranded products seems to assume an importance role in the future. Perhaps, the detersives sold by weight could be explored by leader brands, primarily within the Large Organized Retailers (LRO); 5. the growing specialization of the detersives seems to make to emerge a latent need of simplification in the use and therefore could be explored the role of an eventual detersive specialized for category of need or “universal”.

4.4. The results of empirical survey on consumer goods-Femme Care In reference to the need to feel herself comfort during the menstrual cycle, the analysis makes to emerge a woman much careful to the protection concept of herself and to the health of the products that she acquires. Such emotional involvement is growing among the young women who show more and more sophisticated requirements and needs from this point of view. To the absorbent products it is demanded that they are antiallergenic, therefore clinical tested, transpiring, not annoying, hygienic and above all do not have to be produced with chemical agents and man-made fibers. As the interviews put in evidence, does exist only an element of dissatisfaction for the consumers, referred to the materials by which the sanitary napkin is produced. The customers, in fact, complain a lot for the materials because far from

those natural (like the cotton) or the presence of chemicals fibers that, sometimes, provoke disturbances which irritations and reddening. The features related to the aesthetic and the functional properties, instead, satisfy totally the interviewed (particularly in regard to the brand mainly used LINES, in the version “Seta Ultra con Ali”) for which the attention has to be moved toward those elements of protection of their body that is related to the sphere of personal comfort. Ultimately, the latent need is to have a new product that maintains all the existing functions but using material that are more natural and more neighbors to the intimate sphere of the woman. Regarding, instead, the purchase and use of the products of inner absorbance, it has been emphasized a variable that generate numerous negative perceptions. The purchase of the product, in fact, is conditioned from ties of psycho-cultural nature, regarding: a. to the mentality of individuals; b. the fears connected to problematic use of the product, many times driven by psychological and cultural resistances. Therefore, the core benefit of inner absorbance products (“to feel herself free”) is not sufficient to generate an overcoming of the psychological and cultural barriers. A possible explanation of this phenomenon can be searched in the role of the mother of the young women consumers. For the purchase of this product, the mother does not come consulted, she isn’t more the reference for cultural and psychological related issues. Disappearing this important “mentor”, the consumers, since they are young, don’t take at all in consideration the hypothesis of using the product. The mother accompanies the young woman in the first phase of the event (first menstrual cycle) and, in the majority of the cases, she has an hostility toward the inner absorbents. For many young female consumers, this aspect affects negatively, to a level conscious and unconscious, the relationship with the product. Therefore, even if there isn’t acrimony towards the product, its use comes limited to specific situations.

5. The empirical survey on firm: the findings regarding the case studies The empirical survey on the firms has been carried out applying case study methodology (Yin 2002; Flyvbjerg , 2006), in order to explore how four leading firms, in the specific categories of products and services surveyed (consumer goods pasta, detersives, femme care and electronic consumer goods–mobile phone): a. perceive the role of consumer insight; b. understand the main techniques they use to explore it; c. translate the customer insights in the innovation process, in terms of new products development. In order to develop the case studies, during the period July- Septmber 2009, we have personally interviewed: Irene Larcher, Media and Marketing Services Director of Henkel Italy Spa; Elisabetta Ascione, Consumer Insight Manager of Nokia

Journal of Marketing Trends - Volume I (February 2011) 11


24366_int_Mise en page 1 22/12/10 12:25 Page12

J O U R N A L O F M A R K E T I N G T R E N D S - P R O D U C T I N N O V AT I O N

Italy SpA; Paolo Melegari, Market And People Insight Manager of Barilla Spa; Rita Cantarini, Customer Marketing Knowledge Manager of Fater Spa. In the following we give just a synthetic summary of the main evidences arisen through the interviews. First of all, it emerges a common trait for all case studies: there is a strong tendency to increase the resources and the use of qualitative market research in respect to quantitative ones for the marketing innovation. Fater Spa is leader in the Italian market of the absorbency products for the person (Lines, Tampax, Linidor, Dignity are some leading brands). The marketing qualitative research techniques mainly adopted are the following ones: traditional techniques (focus group and personal experience interviews), to inform the segmentation analysis (the quantitative one); brainstorming on young employee women not operating in the Marketing Department, to stimulate new ideas and perceptions to deepening in quantitative survey; on teenagers, the employment of young women researchers (traditional techniques at individual, group level and also on web by chat); ethnography on the shoppers (shopping visit), to explore purchase dynamics and to evaluate the effects of the promotional strategies and/or co-marketing with the Large Organization Retailers labels; ethnography on the consumers (meeting with consumers), to test the launch of a new product of towels soaked in water for the child hygiene; diaristic (28 days research), to survey consumers perceptions during menstrual cycle and to define a new claim for the sanitary napkins for the night; personal interviews and focus groups, to select the claims for launching new lines of products for the absorbency. Henkel Italia Spa is a subsidiary of Henkel AG & Co. KgaA, that holds globally leading market positions in the business areas: Laundry & Home Care, Cosmetics/ Toiletries and Adhesive Technologies. The marketing qualitative research techniques mainly adopted are the following ones: traditional techniques (focus group and personal experience interviews); concept clinic that tests the “product concepts” during the focus groups modifying in real time, with the collaboration of a designer, the color of the product, the label, the format and recording the reactions of the consumers to the changes; during the interviews and focus groups is used the technique of the metaphors (chosen of the images, associations, collage and interpretation of the image composed through the collage); ethnography on the consumers (home visit), to understand the existing gap between what they declare and what they really make at home day by day; ethnography on the shoppers (shopping visit), to find out, through the observation of the consumer in the point of sale, not satisfied needs.; “skill lab” or competitive focus group: “we put strong Dash users and strong Dixan users in competition each other in order to speed up the contradictions”, as answered us the manager interviewed. An advanced program on neural-

12 ISSN 1961-7798 - © 2009, International Marketing Trends Conference

marketing also exists, at international level. It was emerging, from focus groups, the sensibility of the consumers to the natural ingredients or the grandmother remedies (i.e. like vinegar). Nelsen “Carboni Attivi” and “Dixan Lavanda” are examples of product innovations that have given shining results in such a direction. Barilla Spa is a leader in the pasta business worldwide, for the pasta ready sauces business in continental Europe, for the bakery products business in Italy, and for the crisp bread business in Scandinavia. Marketing research approach of Barilla Spa is strategic. The insight management office of Barilla (called Market & People Insight) has been created in 1997. The qualitative marketing researches in Barilla cover all the phases of new product development process, as following: 1. Product Design Experience (PDE). It is a methodology of designing a new product by the perceptions and the tastes of the consumers, surveyed through focus groups and under supervision by management of the involved functions (Marketing, R&D, Market & People Insight). The “Armonie” and the “Campagnole” of Mulino Bianco, the “Emiliane” of Barilla, are some of the new products developed through the PDE technique; 2. Ethnography (Real Life Experience). It is realized through a team composed by representatives coming from the Marketing, R&D, Sales, Production, also international. The ethnography regards categories of already existing products in the market. The more recurrent modalities are the following ones: a. safari with the consumer in places of interest; b. shopping accompanied with the consumer; c. home visit with observation participated in various moments of consumption (breakfast, lunch, dinner). Some of the product innovations emerged through the application of such technique are: “Le parole della pasta”, “I valori (del Mulino Bianco) ieri e oggi”, “Il Sugo Carbonara”, the “In store bakery lay out” (that is, the analysis of the composition of the products’ bakery assortment offered to the big retail chains); 3. Creativity. In this case, the key organizational functions committed on the customer (Marketing, R&D, Market and People Insight) work together to create and develop new concepts of product or new brands’ positioning. We speak in this case of “Concept lab and/or Get in touch”. One of the innovations of greater success developed by the application of such technique was the “Ready Meals” in France; 4. Web. It’s totally a new emergent approach to listen the vox populi. “We send physically the products to the consumers; the interaction instead proceeds on line”, as said the manager interviewed. One of the innovations of greater success developed by the application of such technique are INCONTRI and SAVOURY of Mulino Bianco. Nokia Italia Spa is a subsidiary of Nokia Corporation, globally leading company in the mobile solutions market. The researches on consumer insight for the product innovation are exclusively of competence of the Market Research unit worldwide. The Market Research Depart-


24366_int_Mise en page 1 22/12/10 12:25 Page13

J O U R N A L O F M A R K E T I N G T R E N D S - P R O D U C T I N N O V AT I O N

ment of Nokia Italy, even in relation to the importance of the market, is frequently involved in the global researches, in particular in the focus groups. This involvement is at tactical level for the creativity: for example, to select between various layout of advertising communication for their understanding or approval, in order to proceed with the quantitative survey to evaluate them. The marketing qualitative research technique mainly adopted are the following ones: traditional techniques (focus group and personal experience interviews); then, generally, it follows a quasi-structured analysis through a questionnaire with a series of “open” questions lead mostly from sociologists because they want in particular to pick many aspects in depth. On global level are diffused the observation and the focus group. Such techniques are very useful because they assure a language more narrow the final consumers and the intermediate customers. “I often find myself to use during the meeting some citations arisen from the focus groups. They concur to be closer to the consumers and customers cause they change our mind set and the shared cognitive scheme of the firm”, as said the manager interviewed.

6. Conclusions, managerial implications and future research directions. Relatively to product categories for which the emotional involvement of the consumer is very strong, and the product technology is quite homogenous among competing brands, the marketing innovation becomes primary source of competitive advantage. Many brands that compete in mature markets are trying to go beyond the search of new functional drivers for innovation, to advantage of those emotional and experiential ones. Therefore, in particular for the leading brands, a fine tuning of the innovative activities becomes a priority, but, more in general terms, of the shared cognitive scheme of the firm, with the lived life, experiences and emotions of the consumers, which demand a deep understanding of their needs. The generation and the testing of new insights therefore assume a primary role that it can lead to radical innovations in the way to conceive the needs of the customer, also of those more unconscious and latent ones, such as to manage the marketing mix variables. The case studies analysed in the paper, in particular Barilla SpA, have evidenced as the qualitative marketing research techniques assume an high importance, and in certain circumstances, are substitutive of the quantitative ones, in new product development process. As the case study Barilla has put in evidence, such techniques, in fact, permeate the process of new product development not only in the early phase - when they demand a more creative or inductive thought on the consumers for the generation of new ideas or insights - but also in much more advanced phase, where they are used for the test of the product concept, and in ultimate ones, where they can

be implemented for the pre and post product launch test. The aim of this paper has been to explore the potentialities of a mix of qualitative research techniques for developing new knowledge and insights on the consumers, presenting the first findings of an ongoing research project. Totally, the field research has involved 12 product-service categories, whose customer insights have been analysed through three complementary qualitative market research techniques (focus group, personal experience narrative-telling story, sentence completion). The design of research includes the analysis of at least one case study for each of twelve product-service category surveyed. In this paper we have reported the findings on the customer insight just on four of twelve categories surveyed; also, they are synthesized the findings about the case studies of four leading firms, one for each category analysed. Therefore, the main limit of such paper is inherent to the peculiarity of the fieldcategories under investigation and the small qualitative empirical evidence on the consumer supported. In the near future the results of the empirical research on the other product-service categories inquired and case studies will be introduced to the attention of the scientific community in the opportune communication forms and channels. In perspective, also, the research design will include for the future the widening of the categories of product-service and the case studies of surveying, as the application of different exploratory techniques (i.e.: ethnography and Zmet) in order to verify value and limits of a possible wider set of such methodologies and to evidence their ability to support the innovation marketing processes of the firms.

References Boddy, C. (2005a), “Projective techniques in market research: valueless subjectivity or insightful reality?”, International Journal of Market Research, Vol. 47, Issue Nr. 3, pp. 239-254. Boddy, C. (2005b), “A rose by any other name may smell as sweet but “group discussion” is not another name for a “focus group” nor should it be”, Qualitative Market Research: An International Journal, Vol. 8, Nr. 3, pp. 248-255. Brannen, J. (ed.)(1992), Mixing Methods : Qualitative and Quantitative Research, London, Gower. Campbell, A.J. (2003), “Creating customer knowledge competence: managing customer relationship management programs strategically”, Industrial Marketing Management, Vol. 32, Nr. 5, pp. 375-383. Chase, S. E. (2005), «Narrative Inquiry: Multiple lenses, approaches, voices», in Denzin, N. K. , Lincoln, Y.S. (EDs), The Sage handbook of qualitative research (3rd ed.), Thousand Oaks, CA, Sage, pp. 651-679.

Journal of Marketing Trends - Volume I (February 2011) 13


24366_int_Mise en page 1 22/12/10 12:25 Page14

J O U R N A L O F M A R K E T I N G T R E N D S - P R O D U C T I N N O V AT I O N

Corbin, J., Strauss A. (1990), “Grounded Theory Research: Procedures, Canons and Evaluative Criteria”, Qualitative Sociology, Vol. 13, Nr. 1. pp. 3–21. Covà, B. (2003), Il marketing tribale. Legame, comunità, autenticità, come valori del marketing mediterraneo, Milano, Il Sole 24 Ore Libri. Davenport, T.H., Harris, J.G., Kohli, A.K. (2001), “How do they know their customers so well?”, MIT Sloan Management Review, Vol. 42(2), pp. 63–73.

Li, T., Calantone, R.J. (1998), «The impact of market knowledge competence on new product advantage: conceptualization and empirical examination», Journal of Marketing, Vol. 62, October, pp.13-29. Lundvall, B.A. (1985), Product Innovation and User-Producer Interaction, Aalborg, Aalborg University Press. Milliken, J. (2001), “Qualitative research and marketing management”, Management Decision, 39 (1): 71-77.

Deshpande, R. (1983), «Paradigms Lost»: On Theory and Method in Research in Marketing”, Journal of Marketing, 47 (Fall): pp. 101-110.

Pinnegar, S., Daynes, J. S. (2007), «Locating narrative inquiry historically. Thematics in the turn to narrative», in Clandinin, D. J. (ed.), Handbook of narrative inquiry: mapping a methodology, Thousand Oaks, CA, Sage Publications, pp. 3-8.

Flyvbjerg, B. (2006), «Five Misunderstandings About Case Study Research», Qualitative Inquiry, Vol. 12, April, Nr. 2, pp. 219-245.

Pralahad, C. K., Hamel, G. (1990), “The core competence of the corporation”, Harvard Business Review, Vol. 68, May-June, pp. 79– 91.

Day, G. S. (1994), “The capabilities of market-driven organizations”, Journal of Marketing, Vol. 58, October, pp. 37– 52.

Gabriel, Y., Griffiths, D.S. (2004), “Stories in organizational research”, in Cassell, C., Symon, G. (Eds.), Essential Guide to Qualitative Methods in Organizational Research, London, Sage, pp. 114-126. Gibbert, M., Leibold, M., Probst, G. (2002), “Five Styles of Customer Knowledge Management, and How Smart Companies Use Them To Create Value”, European Management Journal, Vol. 20, Nr. 5, pp. 459–469. Glazer, R. (1991), “Marketing in an information-intensive environment: strategic implications of knowledge as an asset”, Journal of Marketing, 59, Nr. 55, October, pp.1– 16. Gummesson, E., (2005), “Qualitative Research in Marketing”, European Journal of Marketing, 39 (3–4), pp. 309–311. Hunt, S. D. (1991), “Positivism and Paradigm Dominance in Consumer Research: Toward Critical Pluralism and Rapprochement”, Journal of Consumer Research, Vol. 18, June, pp.32-44.

Riessman, C. K. (2008), Narrative methods for the human sciences, Thousand Oaks, CA, Sage Publications. Saunders, J. (1999), Quantitative methods in marketing, London, Thompson. Sawhney, M. (2004), “Insights into Customer Insights”, CRM Project, Vol. 5, October, pp-2-4. Siegel, H. (1988), “Relativism for consumer research? (Comments on Anderson)”, Journal of Consumer Research, 15 (June), 129-132. Spanjaard, D., Freeman, L. (2006), “Is Qualitative Research always Exploratory?”, Conference 2006, School of Advertising, Marketing and Public Relations, Faculty of Business, Australian and New Zealand Marketing Academy (ANZMAC), 4-6 December, Brisbane Queensland.

Hyde, K. F. (2000), “Recognizing deductive processes in qualitative research”, Qualitative Market Research, 3 (2): pp. 82-89.

Stafford, M. R., Stafford, Th. F. (1993), “Participant Observation and the pursuit of truth: methodological and ethical considerations”, Journal of the Market Research Society 35 (1), pp. 63-77.

Jeppesen, L. B. (2001), “Making Consumer Knowledge Available and Useful. The Case of the Computer Games Industry”, DRUID Nelson and Winter Conference, Summer, Aalborg, Denmark, June 12-15.

Von Hippel, E. (1988), Sources of Innovation, Oxford, Oxford University Press.

Kuhn, T. (1962), The Structure f Scientific Revolutions, Chicago, University of Chicago Press. Laudan, L. (1976), “Two dogmas of methodology”, Philosophy of Science, 55, pp. 25-38. Lawrence, C. Soley, Aaron, L. Smith (2008), Projective techniques for social science and business research, Milwaukee, Wis., Southshore Press.

14 ISSN 1961-7798 - © 2009, International Marketing Trends Conference

Yin, R. K. (2009), Case Study Research. Design and Methods (4th ed.), Thousand Oaks, CA, Sage Publications. Zaltman, G., Zaltman, L. (2008), Marketing Metaphoria: What Deep Metaphors Reveal about the Minds of Consumers, Boston, Harvard Business School Press. Zaltman, G. (2003), How Customers Think: Essential Insights Into the Mind of the Market, Boston, Harvard Business School


24366_int_Mise en page 1 22/12/10 12:25 Page15

JOURNAL OF MARKETING TRENDS - INTERNET

Modellgestützte Präferenzanalyse im Internet als Grundlage für eine kundenorientierte Angebotspolitik Offering Customer-oriented Products on the Internet Using Model-based Preference Analysis Im vorliegenden Beitrag wird ein neues Modell auf Basis der PoissonRegression zur Analyse von Konsumentenpräferenzen im Internet vorgestellt. Die gewählte Vorgehensweise weist Analogien zur Conjoint-Analyse auf und ermöglicht Einblicke in die der Inanspruchnahme bzw. Nichtinanspruchnahme alternativer Produkt- und Serviceangebote zugrunde liegenden Präferenzstrukturen. Im Gegensatz zur Conjoint-Analyse wird dabei auf Dialogdaten zurückgegriffen, wie sie im Zusammenhang mit individuellen Anfragen im Internet automatisch anfallen. Das Modell liefert auf diese Weise Ansatzpunkte zur Verbesserung des gegenwärtigen Angebotsspektrums. Die prinzipielle Vorgehensweise wird am Beispiel von Nutzungsdaten eines kommerziellen Internetportals und Online-Buchungsdaten eines bekannten Reiseveranstalters demonstriert. This paper presents a new model based on Poisson regression for analysing consumer preferences on the Internet. Our approach bears resemblance to conjoint analysis and delivers insights into the preference structures underlying the usage or non-usage of alternative product and service offerings. In contrast to conjoint analysis, the model uses Internet transaction data as it automatically results when a request is launched. In doing so, the model provides useful information with respect to potential improvements of the current range of products or services. The basic idea of our approach is demonstrated by means of usage data provided by a commercial web portal and online booking data provided by a well-known tour operator. Schlüsselwörter: Analyse von Internetangeboten, Log-File-Daten, Neuproduktentwicklung, Poisson-Regression, Präferenzanalyse, Produktplanung Analysis of Internet offerings, log file data, new product development, Poisson regression, preference analysis, product planning

> Reinhold Decker Bielefeld University, Germany rdecker@wiwi.uni-bielefeld.de

> Sören W. Scholz Bielefeld University, Germany

Journal of Marketing Trends - Volume I (February 2011) 15


24366_int_Mise en page 1 22/12/10 12:25 Page16

JOURNAL OF MARKETING TRENDS - INTERNET

1. Motivation und Literatur

1. Motivation and Literature

Die Analyse und das Verständnis des Kauf- und Konsumverhaltens im Internet ist ein zentrales Anliegen der Marketingforschung (Cheung et al. 2003) und eine wichtige Voraussetzung für die Gestaltung von Internetseiten, „die aus Besuchern zahlende Kunden machen“ (Venkatesh und Agarwal 2006, p. 267). Mittlerweile existiert eine unüberschaubare Menge von Beiträgen, die sich mit den Vor- und Nachteilen des Internets als Basis für den Kauf oder Verkauf von Produkten und Dienstleistungen unterschiedlichster Art auseinandersetzen (Steckel et al. 2005). In einzelnen Bereichen, so z. B. im Tourismus oder in der Musikbranche, erfolgt bereits heute ein beträchtlicher Teil der Kaufhandlungen unter Einbeziehung des Internets. So besuchen z. B. gemäß der Power & Associates 2006 New Autoshopper.com Studie 59% der Neuwagenkäufer mindestens eine HerstellerWebsite, die meisten sogar mehrere, bevor sie einen Kauf tägigen (www.jdpower.com). Eine anspruchsvolle Aufgabe stellt in diesem Zusammenhang die Identifikation von Verhaltensmustern dar, auf deren Basis der Frage nachgegangen werden kann, wann und warum es bei einem Internetnutzer zu einer Kaufhandlung kommt (Constantinides 2004; Montgomery et al. 2004). Ein möglicher Ansatz, um solche Verhaltensmuster zu identifizieren, ist die Durchführung von Primärstudien. Erfolgen diese online, so erweist sich die hohe Flexibilität des Mediums Internet bei der Ausgestaltung der Frageinhalte als vorteilhaft, z. B. die Filterführung betreffend (Meißner et al. 2008; Dahan und Srinivasan 2000). Als Nachteil muss hingegen u. a. die Tatsache gewertet werden, dass die Befragten, je nach Thema, zu sozial erwünschten Antworten neigen können. Alternativ besteht die Möglichkeit einer modellgestützten Analyse der bei realen Kundenkontakten im Internet aufgezeichneten Daten, wie sie auch beim Web Usage Mining Verwendung finden (Spiliopoulou 2000). Der vorliegende Beitrag verfolgt diesen zweiten Ansatz und kommt somit ohne eigene Datenerhebung aus. Die Zweckmäßigkeit der sekundärdatengestützten Methodik wird anhand zweier realer Beispiele demonstriert. Das erste Beispiel bezieht sich auf Internetportale. Ein Internetportal bietet seinen Nutzern Leistungen, die sich, wie das in Abbildung 1 dargestellte Beispiel dokumentiert, aus unterschiedlichen Bestandteilen (Rubriken) zusammensetzen. Die unter der Überschrift „TopThemen“ zusammengefassten Rubriken stellen in der nachfolgend beschriebenen Methodik Leistungsmerkmale dar, die ihrerseits wiederum in verschiedenen Ausprägungen vorliegen können und zumeist nominales Skalenniveau aufweisen. Das Merkmal „Auto“ beispielsweise besaß zum Zeitpunkt der Vorbereitung dieses Beitrags u. a. die thematisch sortierten Ausprägungen (Unterrubriken) „Gebrauchtwagen“, „Neuwagen“, „Routenplaner“, „Rallye“, „Finanzierung“ und „Autogalerien“.

The analysis and understanding of purchase and consumption behaviour on the Internet is a major concern for marketing research (Cheung et al. 2003) and an important prerequisite for the construction of websites “that can turn visitors into paying customers” (Venkatesh and Agarwal 2006, p. 267). Meanwhile, there exists a vast amount of literature discussing the advantages and disadvantages of the Internet as a basis for the purchase or sale of products or services of varying kinds (Steckel et al. 2005). In some areas, such as in tourism or in the music industry, a significant part of the buying activities is already carried out on the Internet. For example, according to the Power & Associates 2006 New Autoshopper.com Study 59% of new vehicle buyers visit at least one manufacturer website before making their purchase, but most even visit several sites (www.jdpower.com). A challenging task in this context is the identification of behaviour patterns that are helpful in answering the question of when and why an Internet user makes a purchase (Constantinides 2004; Montgomery et al. 2004). One possible approach to identify such behavioural patterns is to conduct primary studies. If these studies are carried out online, web-based technologies allow highly flexible surveys, e.g. using sophisticated skip patterns in online questionnaires (Meißner et al. 2008; Dahan and Srinivasan 2000). However, there are a few drawbacks when using (online) surveys. One disadvantage is that respondents, depending on the particular topic, may tend to provide socially desirable answers. Alternatively, real customer transaction data recorded on the Internet (as also employed in web usage mining (Spiliopoulou 2000)) can be used to analyse and to model behavioural patterns. This paper follows the latter approach and consequently gets along without collecting primary data. The appropriateness of the methodology based on online secondary data is demonstrated by means of two real world examples. The first example refers to Internet portals. An Internet portal offers its users contents that are – as shown in Figure 1 – composed of different constituent parts (content categories). In the subsequently described methodology, the categories that are combined under the heading “Top-Themen” (top topics) represent attributes that are available in various realisations (called levels in the following) and are mostly nominally scaled. When preparing this paper, the attribute “Auto” (motor car), for example, comprised various levels (sub-categories) according to the topics “Gebrauchtwagen” (used cars), “Neuwagen” (new cars), “Routenplaner“ (route planner), “Rallye” (rally), “Finanzierung” (financing), and “Autogalerien” (auto galleries).

16 ISSN 1961-7798 - © 2009, International Marketing Trends Conference


24366_int_Mise en page 1 22/12/10 12:25 Page17

JOURNAL OF MARKETING TRENDS - INTERNET

Abbildung 1: Möglicher Aufbau eines Internetportals

Neben den unterschiedlichen Rubriken enthalten die einzelnen Seiten eines Internetportals in der Regel auch Werbeanzeigen (z. B. in Form von Bannern), deren Wirkung umso stärker zum Tragen kommen kann, je länger sich die Besucher des Portals auf der betreffenden Web-Seite aufhalten. Die Aufenthaltsdauer auf den Seiten eines Internetportals (z. B. gemessen in Minuten) kann somit als Maß für seine Attraktivität aus Nutzersicht interpretiert werden und stellt im Folgenden die metrisch skalierte Responsevariable dar. Diese Vorgehensweise ist in der Web Usage-Analyse weit verbreitet (Burklen et al. 2005). Tao et al. (2006) diskutieren in diesem Zusammenhang die Problematik, dass in Log-Files im Normalfall nur der Aufruf und das Verlassen einer Seite gemessen werden. Die Attraktivität eines Internetportals resultiert im Wesentlichen aus dem Nutzen der angebotenen Inhalte und/oder Services. Gleichzeitig wird ein Portal für auf diesem Wege Werbung treibende Unternehmen umso interessanter, je höher die Aufenthaltsdauer potenzieller Kunden auf den betreffenden Seiten ausfällt, womit sich die Kausalkette schließt. Dies führt zu der Frage, welchen Beitrag die einzelnen Rubriken zur Gesamtattraktivität des Portals liefern. Zuverlässige Erkenntnisse hierüber können dazu beitragen, bestehende Portale noch kundengerechter zu gestalten. Als Analysegrundlage bieten sich die in Log-Files aufgezeichneten Portalnutzungsdaten an. Welche technischen Probleme sich im Zusammenhang mit der analysegerechten Aufbereitung von Log-File-Daten ergeben können, wird z. B. bereits bei Cooley et al. (1999) thematisiert und ist nicht Gegenstand der weiteren Ausführungen. Das zweite Beispiel entstammt dem Bereich Tourismus. Nahezu alle bedeutenden Touristikunternehmen bieten ihre Produkte heute auch oder sogar ausschließlich im Internet an. Die Kunden haben hier die Möglichkeit, sich eigenständig über alternative Reiseangebote zu informieren und diese online zu buchen (Kuom und Oertel 1999; Law et al. 2004). Der anhaltende Trend zu OnlineReisebuchungen wird vor allem durch die schnellen Reservierungs- und Buchungsmöglichkeiten, die Bündelung von aktuellen Informationen und Angeboten zu bestimmten Reisegebieten und die Reduktion der Kosten durch

Figure 1: Possible Structure of an Internet Portal

In addition to the different categories, the individual pages of an Internet portal usually also contain advertisements (e.g., in the form of banners). The effects of these advertisements can be the stronger the longer a visitor stays on the respective web site. The length of stay on the pages of an Internet portal (e.g., measured in minutes) can be interpreted as a measure of its attractiveness from a user perspective and is represented in the following by a metric response variable. This approach is widespread in web usage analysis (Burklen et al. 2005). In this context, Tao et al. (2006) discuss some of the difficulties associated with log files, e.g., the fact that normally only the call and leaving of a web page are measured. The attractiveness of an Internet portal is due mainly to the benefits of the contents and/or services offered. At the same time, a portal becomes the more interesting for companies that carry out online advertisement the longer potential customers stay on the respective web pages, which completes the causal chain. This leads to the question to which extent the individual content categories contribute to the overall attractiveness of an Internet portal. Reliable insights about this can support the design of even more customer-friendly portals. Log file data provides the basis for such analyses. Technical problems that can arise in connection with the preparation and analysis of log file data are already discussed, for example, in Cooley et al. (1999) and are not in the focus of this paper. The second example is taken from the field of tourism. Almost all major tour operators sell their products on the Internet today. Some of them even use this distribution channel exclusively. Here, the customers have the opportunity to acquire information about alternative travel offers and to book them online (Kuom and Oertel 1999; Law et al. 2004). The continuing trend towards online travel bookings is primarily increased by fast reservation and booking facilities, the pooling of current information and offers regarding certain travel areas, and the reduction of costs through efficient internal operations (Oertel et al. 2003). The starting point of the subsequent discussions is transactional data as it arises from a dialogue like the one indicated by Figure 2.

Journal of Marketing Trends - Volume I (February 2011) 17


24366_int_Mise en page 1 22/12/10 12:25 Page18

JOURNAL OF MARKETING TRENDS - INTERNET

effiziente interne Operationen verstärkt (Oertel et al. 2003). Ausgangspunkt der nachfolgenden Überlegungen sind Transaktionsdaten, wie sie z. B. bei einem Dialog der durch Abbildung 2 angedeuteten Art entstehen können.

Figure 2: Elements of a Product Specification with an Online Tour Operator

Abbildung 2: Elemente einer Leistungsspezifikation bei einem Online-Reisebüro

Hierbei stellen der Abflugort und das Reiseziel Merkmale der Leistung „vermittelte Reise“ dar. Mögliche Ausprägungen der genannten Merkmale sind z. B. „Düsseldorf“ bzw. „Palma de Mallorca“. In beiden Fällen handelt es sich um nominale Merkmale, während das aus Hin- und Rückflugdatum resultierende Merkmal Reisedauer (in Tagen) als metrisch interpretierbar ist. Weitere Merkmale können aus der Hotelreservierung sowie zusätzlich angebotenen Services wie Mietwagen und Reiserücktrittsversicherung resultieren. Besucht eine Person die Web-Seiten eines Reiseveranstalters, so kann dies die folgenden Gründe haben: 1. Sie hat die konkrete Absicht, auf diesem Wege eine Reise zu buchen („Buchung“). 2. Sie möchte sich ohne konkrete Buchungsabsicht über die vom kontaktierten Reiseveranstalter angebotenen touristischen Offerten informieren („Anfrage“). 3. Sie landet aus nicht mit einem konkreten Reiseinteresse verbundenen Gründen auf den Web-Seiten. Der zuletzt genannte Fall kann bei den weiteren Betrachtungen außer Acht bleiben. Die beiden anderen Punkte werfen hingegen die unter Produktgestaltungsgesichtspunkten interessante Frage auf, welche im Dialog spezifizierten Kombinationen von Merkmalsausprägungen häufig zu einer Online-Buchung führen und welche eher in einer der Informationsbedarfsbefriedigung dienenden Anfrage ihren Niederschlag finden. Die Beantwortung dieser Frage liefert nicht nur Hinweise auf die Attraktivität der gegenwärtig angebotenen Leistungen, sondern sie kann auch Anhaltspunkte für die Ausgestaltung der zukünftig bereitzustellenden Leistungen geben. Insoweit weist die in Abschnitt 2 vorgestellte Methodik Analogien zur Conjoint-Analyse auf (Backhaus et al. 2007; Baier und Brusch 2009), kommt aber im Gegensatz zu dieser ohne eigens durchzuführende, oftmals kostspielige Datenerhebung aus und ist nicht-linear in den Parametern. Bei Conjoint-Analysen stellt darüber

18 ISSN 1961-7798 - © 2009, International Marketing Trends Conference

In this example, departure place and travel destination represent attributes of the product/service “package tour”. Possible levels of the aforementioned attributes are, for example, “Düsseldorf” and “Palma de Mallorca”. In both cases, the attributes are nominal. On the contrary, the attribute “duration of trip” (in days), as the difference of outbound and return flight date, can be interpreted as metric. Further attributes can result from hotel reservation as well as from additional services offered, like car rental and travel cancellation insurance. If a person visits the web pages of a tour operator, this may be due to the following reasons: 1. He or she has a clear intention to book a trip (“booking”). 2. He or she wants to get information about the touristic offers of the contacted tour operator without a clear booking intention (“inquiry”). 3. He or she visits the web pages without any concrete interests related to travelling. The last case can be disregarded and will no longer be considered in the following. The other two reasons, however, are raising the following question concerning future product designs: Which attribute levels increase the number of bookings and which ones rather increase the number of visits of persons who just want to satisfy their information needs? Answering this question does not only provide information on the attractiveness of the services currently offered but can also stimulate new product development by delivering clues for the design of future services. In this regard, the methodology introduced in Section 2 bears resemblance to conjoint analysis (Backhaus et al. 2007; Baier and Brusch 2009). But in contrast to this, it can be implemented without carrying out a special and often costly data collection, and it is non-linear in the parameters. Furthermore, the limited number of attributes and attribute levels that can be considered in conjoint analyses is an often-cited problem (Meißner et al. 2007; Scholz et al. 2009).


24366_int_Mise en page 1 22/12/10 12:25 Page19

JOURNAL OF MARKETING TRENDS - INTERNET

hinaus die Anzahl einbeziehbarer Merkmale und Merkmalsausprägungen eine immer wieder thematisierte Problematik dar (Meißner et al. 2007; Scholz et al. 2009).

2. Preference Analysis using Poisson Regression

2. Präferenzanalyse mittels Poisson-Regression

The following model is based on three key assumptions: 1) The products or services to be analysed represent combinations of levels of a fixed number of attributes. 2a) The dwell time on an Internet portal can be used as an indicator of its attractiveness. 2b) The completion of an online booking and the resulting purchase act substantiate the basic acceptance of the respective offer. The notation used to describe the structure of the analyzed data and to specify the following models is summarized in Figure 3. In addition, Table 1 illustrates the data base underlying the empirical study.

2.1 Annahmen und Datenstruktur Die nachfolgende Modellierung basiert auf drei zentralen Annahmen: 1) Die zu analysierenden Produkte oder Services stellen Kombinationen von Ausprägungen einer festen Anzahl von Leistungsmerkmalen dar. 2a) Die Aufenthaltsdauer in einem Internetportal kann als Maß für dessen Attraktivität verwendet werden. 2b) Das Zustandekommen einer Online-Buchung resp. der damit verbundene Kaufakt dokumentieren die grundsätzliche Akzeptanz des betreffenden Angebots. Die zur Beschreibung der Struktur der in die Analyse eingehenden Daten und zur Spezifikation der nachfolgenden Modelle verwendete Notation ist in Abbildung 3 dargestellt. Tabelle 1 illustriert darüber hinaus die der empirischen Studie zugrundeliegende Datenbasis.

2.1 Assumptions and Data Structure

Figure 3: Notation Used in the Poisson Regression Framework

Abbildung 3: In den Poisson-Regressionsmodellen verwendete Notation

The combinations of the L product or service attributes with the response variables “inquiry status” and “booking status” describe the data base illustrated by means of the tourism example (see Figure 2). Hereafter, in specification 2, attribute 1 was selected with level 1 whereas attribute L was selected with level 3. The respective profile led to a booking (yk2 = 1).

Die Kombinationen der L Leistungsmerkmale mit den Responsevariablen „Anfragestatus“ und „Buchungsstatus“ beschreiben die durch das Tourismusbeispiel veranschaulichte Datengrundlage (vgl. Abbildung 2). Hiernach wurde bei Spezifikation 2 das Merkmal 1 in Ausprägung 1, das Merkmal L hingegen in der Ausprägung 3 gewählt. Das entsprechende Leistungsprofil führte zu einer Buchung (yk2 = 1). Tabelle 1: Struktur der disaggregierten Datenbasis

In analoger Weise spezifiziert eine Kombination der L Leistungsmerkmale mit der Responsevariable „Konsultationsdauer“ die durch das Portalbeispiel veranschaulichte

Table 1: Structure of the Disaggregated Data Base

In a similar manner, a combination of the L product or service attributes with response variable “duration of consultation” specifies the data base which was illustrated using the portal example (see Figure 1). In this reading, however, it is quite possible that several levels of an attribute take the value 1 at the same time. If all levels of all attributes take the value 0, this indicates that the respective consultation was started via the homepage of the portal, but no further entry into the portal followed. The respective visits are not considered in the analysis. Specification 1 in Table 1 can thus be interpreted as follows: Within this consultation, subcategory (level) M1 of category 1 and subcategory (level) 2 of category L were

Journal of Marketing Trends - Volume I (February 2011) 19


24366_int_Mise en page 1 22/12/10 12:25 Page20

JOURNAL OF MARKETING TRENDS - INTERNET

Datengrundlage (vgl. Abbildung 1). In dieser Lesart ist es allerdings durchaus möglich, dass mehrere Ausprägungen eines Merkmals gleichzeitig den Wert 1 annehmen. Besitzen alle Ausprägungen aller Merkmale den Wert 0, so bedeutet dies, dass bei der betreffenden Konsultation zwar auf die Eingangsseite des Portals zugegriffen, auf einen weiteren Einstieg in das Portal aber verzichtet wurde. Die betreffenden Datensätze bleiben bei der Analyse unberücksichtigt. Spezifikation 1 in Tabelle 1 wäre somit wie folgt zu interpretieren: Im Rahmen dieser Konsultation wurde in Rubrik 1 auf die Unterrubrik M1 und in Rubrik L auf die Unterrubrik 2 zugegriffen. Die Zugriffe führten in Summe zu einer Gesamtaufenthaltsdauer von drei Minuten. Die durch Tabelle 1 veranschaulichten Rohdaten können nun in der Weise aggregiert werden, dass im Falle des Tourismusbeispiels die Spezifikationen identischer Leistungen zusammengefasst und mit den entsprechenden Häufigkeiten für erfolgte Anfragen und Buchungen versehen werden. Mit I = M1∙M2∙,…,∙ML als der Anzahl möglicher Kombinationen von Merkmalsausprägungen resp. Leistungen und zi1 bzw. zi2 als die auf Spezifikation i entfallende Anzahl an Anfragen bzw. Buchungen ergibt sich die in Tabelle 2 dargestellte Datenstruktur. Auf Spezifikation 2 entfallen demzufolge 34 Anfragen und 5 Buchungen.

Tabelle 2: Struktur der aggregierten Tourismusdaten

Im Falle der Auswertung von Portalnutzungsdaten ist eine solche Aggregation nicht erforderlich. Hier liegen mit Responsevariable yk3 bereits die für eine Modellanwendung nutzbaren Daten vor.

2.2

accessed. All in all, the respective visits resulted in a total dwell time of three minutes. The raw data illustrated in Table 1 can now be aggregated in such a way that, in the case of the tourism example, specifications of identical attribute levels are combined together and furnished with the corresponding frequencies of completed inquiries and bookings. Let I = M1∙M2∙...∙ML be the number of possible combinations of attribute levels and zil and zi2 be the number of inquiries and bookings respectively concerning specification i, then the data structure presented in Table 2 results. Accordingly, 34 inquiries and 5 bookings are allotted to specification 2. Table 2: Structure of the Aggregated Tourism Data

When analyzing portal usage data, such an aggregation is not required. Here, the data needed for applying the model is already given with response variable yk3.

2.2 Analysis of Web Portal Usage If the length of stay on a web portal (e.g., in minutes) is interpreted as the result of a counting process, Poisson regression provides an appropriate framework for modelling the random variable Y3 (with realisation y3) representing the count data (Long and Freese 2006; Drèze and Zufryden 1997, 1998). Accordingly, we specify the probability function of the Poisson distribution given in Figure 4 and implement the indicated reparameterization for λ. Figure 4: Model Used for Web Portal Usage Analysis

Analyse von Internetportalnutzungen

Interpretiert man die Dauer des Aufenthalts im Internetportal (z. B. in Minuten) als Ergebnis eines Zählprozesses, so bietet sich für die Modellierung der durch die Zufallsvariable Y3 (mit Ausprägung y3) repräsentierten Zähldaten die Poisson-Regression an (Long und Freese 2006; Drèze und Zufryden 1997, 1998). Dementsprechend formulieren wir die in Abbildung 4 dargestellte Wahrscheinlichkeitsfunktion der Poisson-Verteilung und nehmen für λ die angegebene Reparametrisierung vor. Aufgrund des dichotomen Charakters der Leistungsmerkmale und mit Bezug auf die in Unterabschnitt 2.1 dargelegte Interpretation der Responsevariable yk3 können die resultierenden Parameterschätzer als mit den Teilnutzenwerten aus der Conjoint-Analyse vergleichbare

20 ISSN 1961-7798 - © 2009, International Marketing Trends Conference

Because of the dichotomous character of the attributes and with respect to the interpretation of response variable yk3 outlined in Subsection 2.1, the


24366_int_Mise en page 1 22/12/10 12:25 Page21

JOURNAL OF MARKETING TRENDS - INTERNET Größen aufgefasst werden. Der Wert des β21-Schätzers entspricht beispielsweise dem Beitrag, den die Ausprägung 1 des Merkmals 2 zu dem über die Konsultationsdauer operationalisierten Gesamtnutzen des Portals leistet. Abbildung 4: Für die Analyse von Internetportalnutzungen verwendetes Modell

2.3 Analyse des Buchungsverhaltens Analog zu den Ausführungen in Unterabschnitt 2.2 und mit Bezugnahme auf die durch Tabelle 2 veranschaulichte Datengrundlage lassen sich auch für das Tourismusbeispiel die Abhängigkeitsbeziehungen zwischen den betrachteten Leistungsmerkmalen und den Responsevariablen „Anzahl Anfragen“ (Zufallsvariable Z1 mit Ausprägung z1) und „Anzahl Buchungen“ (Zufallsvariable Z2 mit Ausprägung z2) unter Rückgriff auf eine Poisson-Regression modellieren. Um dabei explizit zum Ausdruck zu bringen, dass eine Spezifikation entweder in einer Anfrage oder in einer Buchung mündet, bietet sich die kombinierte Erfassung beider Aspekte mittels bivariater Poisson-Regression an. Die entsprechende Wahrscheinlichkeitsfunktion sowie die zur Parameterschätzung zu maximierende Log-LikelihoodFunktion sind in Abbildung 5 wiedergegeben. Analog zum univariaten Fall wird mit den β-Schätzern der Einfluss der m1-ten Ausprägung des Merkmals l auf die über die Anzahl der Anfragen (j = 1) bzw. die Anzahl der Buchungen (j = 2) operationalisierte Gesamtattraktivität der betreffenden Leistung zum Ausdruck gebracht. Die bivariate Modellierung ermöglicht in intuitiver Weise die simultane Optimierung des Modells hinsichtlich beider Kundenreaktionen. Abbildung 5: Für die Analyse des Buchungsverhaltens verwendetes Modell

resulting parameter estimators can be interpreted in a similar way as the part-worth utilities known from conjoint analysis. The value of the β21-estimator, for example, corresponds to the contribution of level 1 of attribute 2 to the total utility of the web portal operationalized by the duration of consultation.

Analysis of Booking Behaviour Analogous to the explanations in Subsection 2.2 and by referring to the data example illustrated by Table 2, the dependencies between the observed attributes and the response variables “number of inquiries” (random variable Z1 with realisation z1) and “number of bookings” (random variable Z2 with realisation z2) is modelled in the case of the tourism example using a Poisson regression as well. In order to explicitly express that a specification either leads to an inquiry or a booking, a bivariate Poisson regression providing a combined coverage of both aspects can be used. The respective probability function and the corresponding log-likelihood function (to be maximized) used to estimate the unknown model parameters are presented in Figure 5. Figure 5: Model used for Booking Behaviour Analysis

Like in the univariate case, the β-estimators quantify the influence of the ml-th level of attribute l on the overall attractiveness of the product or service which is operationalized by the number of inquiries (j = 1) and the number of bookings (j = 2). The bivariate modelling allows in an intuitive manner the simultaneous optimization of the model with respect to both customer reactions here.

Empirical Applications 3.1 Analysis of Log File Data from the Tourism Industry For the following application, transaction data of a well-known online package tour operator could thankfully be used. Altogether, the data set includes 3561 tour bookings each of them represented by 10 attributes featuring different numbers of levels (see Table 3 for details).

Journal of Marketing Trends - Volume I (February 2011) 21


24366_int_Mise en page 1 22/12/10 12:25 Page22

JOURNAL OF MARKETING TRENDS - INTERNET

3. Empirische Anwendungen

Table 3: Attributes and Attribute Levels of Travel Bookings

3.1 Analyse von Log-File-Daten aus der Tourismusbranche Für die nachfolgende Anwendung konnte dankenswerterweise auf Transaktionsdaten eines bekannten Online-Reiseveranstalters zurückgegriffen werden. Der Datensatz umfasst 3561 Reisebuchungen, die jeweils anhand von 10 Merkmalen beschrieben werden. Die hierbei möglichen Ausprägungen sind Tabelle 3 zu entnehmen. Tabelle 3: Merkmale und mögliche Ausprägungen von Reisebuchungen

In die Analyse gingen insgesamt 430 gemäß Tabelle 2 kodierte Spezifikationen (Buchungsprofile) ein. Da eine eindeutige Identifikation der getätigten Anfragen auf Basis der verfügbaren Log-File-Daten nicht möglich war, konnte nur eine univariate Poisson-Regression durchgeführt werden (für ein bivariates Beispiel sei auf Decker und Scholz (2008) verwiesen). Das entsprechende Modell erzielte einen zufriedenstellenden AIC-Wert (Akaike Information Criterion) von -1825,97. Durch eine zusätzliche A-priori-Segmentierung konnte dieser Wert noch deutlich verringert werden, was auf eine gewisse Heterogenität in den Präferenzen der Kunden schließen lässt. Aus Platzgründen sind in Abbildung 6 nur die analog zur Conjoint-Analyse als Teilnutzenwerte interpretierten Parameter der Merkmalsausprägungen für die Gesamtstichprobe dargestellt. Die Parameter wurden so normiert, dass die Summe der Teilnutzenwerte der Ausprägungen eines jeden Merkmals gleich Null ist. Dies entspricht üblichen Darstellungen in der Conjoint-Analyse und erlaubt eine direkte Interpretation des Einflusses einzelner Parameter auf die Buchungshäufigkeit (Kuhfeld 2003).

In this analysis, a total of 430 specifications (i.e. booking profiles), coded according to Table 2, were considered. Since a clear identification of the inquiries made was not possible on the basis of the available log file data, only a univariate Poisson regression was carried out (for a bivariate example we refer the reader to Decker und Scholz (2008)). The respective model achieved a satisfying AIC (Akaike Information Criterion) value of 1825.97. Through additional a priori segmentation, this value could be reduced significantly suggesting a certain heterogeneity in customer preferences. For lack of space, Figure 6 only shows the parameters of the attribute levels for the total sample which, analogous to conjoint analysis, are interpreted as partworth utilities. The estimators were normalized so that the sum of the part-worth utilities of the levels of each attribute equals zero. This corresponds to usual representations in conjoint analysis and allows a straight interpretation of the influence of individual parameters on the booking frequency (Kuhfeld 2003). Figure 6: Normalized Parameter Estimators Based on Aggregate Booking Data

Abbildung 6: Normierte Parameterschätzer auf Basis aggregierter Buchungsdaten

Wald tests were carried out to verify the significance of the relationship between the individual explanatory variables and the dependent variable. The tests show that for any attribute at least two levels have a significant relationship (p < 0.05) to booking frequency. Conversely, this means that each attribute contributes to the explanation of the booking frequencies. The present results reveal that in particular the attribute “accommodation” seems to have a significant

22 ISSN 1961-7798 - © 2009, International Marketing Trends Conference


24366_int_Mise en page 1 22/12/10 12:25 Page23

JOURNAL OF MARKETING TRENDS - INTERNET

Zur Überprüfung der Signifikanz des Zusammenhangs zwischen den einzelnen erklärenden Variablen und der abhängigen Variable durchgeführte Wald-Tests zeigen, dass bei jedem Merkmal mindestens zwei Ausprägungen in einer signifikanten Beziehung (p < 0,05) zur Buchungshäufigkeit stehen. Im Umkehrschluss bedeutet dies, dass jedes Merkmal zur Erklärung der Buchungshäufigkeiten beiträgt. Die vorliegenden Ergebnisse offenbaren, dass insbesondere das Merkmal „Unterkunft“ einen deutlichen Einfluss auf die Buchungshäufigkeit zu haben scheint. Auf aggregierter Betrachtungsebene werden Hotels mit bis zu drei Sternen bevorzugt, was zumindest teilweise mit dem Last-Minute-Charakter des betrachteten Reiseangebots zusammenhängen dürfte. Beim Merkmal „Entfernung zur Stadt“ zeigt sich, dass die Unterbringung in urbanen Gebieten (Entfernung < 1 km) den höchsten Nutzen spendet. Den zweithöchsten Nutzen generiert die Ausprägung „> 6 km“. Die beiden anderen Ausprägungen, d. h. die Entfernungen 1-3 km und 3-6 km, liefern hingegen niedrigere Parameterwerte. Dies lässt die Vermutung zu, dass vor allem Unterkünfte, die nahe am Stadtzentrum liegen, und solche, die sich deutlich abseits von dicht besiedelten Gebieten befinden, bevorzugt werden. Unterkünfte, die eher in den Vorstädten oder Randgebieten (und somit in Mischzonen) liegen, werden tendenziell weniger bevorzugt. Dieses recht markante Präferenzmuster lässt sich in einem gewissen Maße damit erklären, dass unterhaltungsorientierte Urlauber eher die Nähe eines pulsierenden Stadt- und Nachtlebens suchen, während an beschaulicher Erholung interessierte Urlauber eher das ruhige Hinterland bevorzugen. Bezieht man zusätzlich das Interzept (1,735) in die Betrachtungen mit ein, so lässt sich für jedes aus den vorhandenen Merkmalsausprägungen generierbare Reiseprofil die potenzielle Buchungshäufigkeit prognostizieren. Auf diese Weise können auch für neuartige Reiseprofile Prognosen der zu erwartenden Buchungshäufigkeit abgegeben werden. Darüber hinaus kann mittels Simulation überprüft werden, inwieweit Veränderungen bestehender Reiseangebote zu höheren oder niedrigeren Buchungsfrequenzen führen würden. Da Zimmerkontingente häufig im Voraus gebucht werden müssen, kann der Reiseveranstalter durch die vorgeschlagene Methodik die zu erwartenden Buchungshäufigkeiten besser planen und somit Negativeffekte infolge von Überbuchungen reduzieren. Aufgrund der in Abbildung 6 dargestellten Parameterwerte bietet es sich für den Reiseveranstalter an, eher einfache Hotels mit Doppelzimmer und/oder Apartments anzubieten. Des Weiteren sollte die Auswahl geeigneter Partnerhotels eher durch das Merkmal „Entfernung zum Strand“ als durch das Merkmal „Entfernung zur Innenstadt“ beeinflusst werden. Die entsprechenden Parameterwerte zeigen, dass sich eine größere Entfernung zum Strand stärker negativ auf die Buchungshäufigkeit auswirkt als eine

impact on the booking frequency. On the aggregate level, hotels with up to three stars are preferred, which at least partly might be related with the last-minute nature of the considered travel offers. With respect to attribute “distance to downtown”, the resulting parameters indicate that housing in urban areas (distance < 1 km) provides the highest benefit. The attribute level “> 6 km” generates the second highest benefit. The other two attribute levels, i.e., the distances 1-3 km and 4-6 km, however, provide lower parameters. This allows the guess that particularly those accommodations which are located close to the city centre and those which are much far away from densely populated areas are preferred. Accommodations that are located in the suburbs or peripheral areas (and hence in mixed zones) tend to be less preferred. This quite remarkable preference pattern can be explained to a certain degree by the fact that entertainment-oriented tourists opt for the closeness of a pulsating city and night life, whereas those tourists who are interested in tranquil recreation prefer the hinterland. If one additionally considers the intercept (1.735), the potential booking frequency of each travel profile that can be generated with the existing attribute levels can be predicted. In this way, predictions of the expected booking frequency for novel travel specifications are possible as well. Moreover, by means of simulation, the extent to which modifications of existing travel offers would lead to higher or lower booking frequencies can be verified. Since room allotments often have to be booked in advance, the tour operator can better assess the expected booking frequency by using the suggested methodology and thus reduce the negative effects of over-bookings. According to the parameter values presented in Figure 6, it seems advisable for the package tour operator to offer simple hotels with double rooms and/or apartments. In addition, the selection of appropriate partner hotels should be influenced by the attribute “distance to beach” rather than by the attribute “distance to downtown”. The corresponding parameter estimators show that a longer distance to the beach has a more negative impact on the booking frequency than a longer distance to downtown. Furthermore, medium-length trips, i.e., approximately 1-2 weeks, promise a greater acceptance by the customers.

Analysis of Log File Data from a Web Portal A web portal mostly provides a broad choice of services from several categories. The categories are motivated either by content (the so-called content area) or include services where the user interacts with the portal (e.g., when using route planners or online games). The range of services offered by a web portal is usually built hierarchically and can typically be divided into three levels. The first level equals the homepage, where the

Journal of Marketing Trends - Volume I (February 2011) 23


24366_int_Mise en page 1 22/12/10 12:25 Page24

JOURNAL OF MARKETING TRENDS - INTERNET

größere Entfernung zum Stadtzentrum. Zudem versprechen Reisen mittlerer Dauer, d. h. ca. 1-2 Wochen, eine größere Akzeptanz seitens der Kunden.

3.2 Analyse von Log-File-Daten aus einem Internetportal Ein Internetportal stellt zumeist eine sehr breit gefächerte, d. h. aus mehreren Rubriken bestehende Leistung dar. Die Rubriken sind entweder inhaltlich motiviert (der so genannte Content-Bereich) oder umfassen Dienste, bei denen der Nutzer in unterschiedlichem Maß mit dem Portal in Interaktion tritt (z. B. bei der Nutzung von Routenplanern oder Online-Spielen). Das Leistungsangebot eines Internetportals ist in der Regel hierarchisch aufgebaut und lässt sich typischerweise in drei Ebenen einteilen. Auf der ersten Ebene befindet sich die Startseite („Homepage“) des Portals, auf der die Besucher einen Überblick über das Leistungsangebot erhalten und von der aus sie sich zu den interessierenden Rubriken durchklicken können. Die zweite Ebene enthält die zumeist thematisch gruppierten Inhaltsübersichten zu den einzelnen Rubriken, die Unterrubriken. Auf der dritten Ebene findet der Besucher die eigentlichen Inhalte, die ihrerseits, je nach Komplexität, wiederum unterschiedlich stark untergliedert sein können. Für die Überprüfung unserer Methodik standen uns dankenswerterweise Log-File-Daten eines führenden Internetportalbetreibers zur Verfügung. Die Daten repräsentieren 10922 Sessions. Zum Zeitpunkt der Datengenerierung gehörten „Reisen“, „Unterhaltung“, „Erotik“, „Finanzen“, „Musik“ und „Sport“ zu den stark frequentierten Rubriken (vgl. Tabelle 4). Diese sechs Rubriken gingen als binär kodierte Merkmale mit unterschiedlichen Anzahlen von Ausprägungen (Unterrubriken) in das Modell ein. Die übrigen, weniger häufig frequentierten Rubriken wurden in einer synthetischen Rubrik „Sonstige“ zusammengefasst und wie Unterrubriken derselben behandelt. Zu den im Betrachtungszeitraum geringer frequentierten Rubriken zählen z. B. „Karriere/Ausbildung“, „Immobilien“ und „Partnersuche“. Tabelle 4: In die Analyse einbezogene Rubriken und Unterrubriken des Internetportals

24 ISSN 1961-7798 - © 2009, International Marketing Trends Conference

visitors obtain an overview of the contents and services offered and from where they can click through to the categories of interest. The second level usually comprises thematically grouped content overviews of the various categories, the subcategories. At the third level, the visitor finds the particular contents, which, depending on their complexity, can in turn be subdivided again. For the verification of our methodology, log file data from a large German Internet portal operator were kindly made available to us. The available data set represents 10922 sessions. At the time of data collection “travel”, “entertainment”, “eroticism”, “finances”, “music” and “sport” belonged to the most frequented categories (see Table 4). These six categories were considered in the model as binary coded attributes with different numbers of levels (subcategories). The remaining, less frequented categories were combined to a synthetic category “others” and treated like subcategories of it. “Career/ education”, “real estate” and “partner search” belong to the less frequented categories within the relevant period. Table 4: Categories and Subcategories of the Internet Portal Included in the Analysis

In order not to timely censor the identified sessions at the beginning and at the end of the relevant time window (Blossfeld and Rohwer 1995), a buffer of 30 minutes was included both before and after this time interval. Sessions that wholly or partly fell in these buffer zones were excluded from the analysis. Also those sessions in which only the starting page of the portal was visited, were neglected due to their low explanatory power for the focus of the present analysis. After cleaning the data, a total of 10912 complete sessions were available for further analyses. Figure 7 shows selected normalized parameter estimators for individual attribute levels. Values greater than zero indicate – on average – a positive influence of the respective attribute levels on the dwell time, whereas negative values represent the opposite effect. One recognizes that in particular subcategories from the areas


24366_int_Mise en page 1 22/12/10 12:25 Page25

JOURNAL OF MARKETING TRENDS - INTERNET

Um die am Anfang und am Ende des betrachteten Zeitfensters identifizierten Sessions nicht zeitlich zu zensieren (Blossfeld und Rohwer 1995), wurde vor und nach diesem Zeitintervall jeweils ein Puffer von 30 Minuten eingerichtet. Sessions, die ganz oder teilweise in diese „Pufferbereiche“ fielen, wurden von der Analyse ausgeschlossen. Auch jene Sessions, bei denen nur die Startseite des Portals besucht wurde, blieben aufgrund ihrer geringen Aussagekraft für den vorliegenden Analysefokus außer Acht. Nach Bereinigung der Daten standen insgesamt 10912 vollständige Sessions für die weitere Analyse zur Verfügung. In Abbildung 7 sind ausgewählte, normierte Parameterschätzer der einzelnen Merkmalsausprägungen dargestellt. Werte größer Null zeigen einen – im Mittel – positiven Einfluss der betreffenden Merkmalsausprägung auf die Verweildauer an, während negative Werte auf einen gegenteiligen Effekt hindeuten. Man erkennt, dass Unterrubriken aus den Bereichen „Erotik“, „Finanzen“ und „Musik“ zu längeren Verweildauern Anlass gaben, ihren Besuchern also offensichtlich einen vergleichsweise hohen Nutzen spendeten. Abbildung 7: Normierte Parameterschätzer für das Internetportal

Der Mehrwert der modellbasierten Analyse wird deutlich, wenn man die in Abbildung 7 dargestellten Parameterschätzer den in Tabelle 4 angegebenen Besuchsfrequenzen je Rubrik bzw. Unterrubrik gegenüberstellt. Die Rubrik „Immobilien“ weist z. B. nur eine vergleichsweise geringe Anzahl an Besuchen auf und zählt somit zu den weniger häufig genutzten Inhalten des Portals. Gleichzeitig kommt aber durch den zugehörigen Parameterschätzer 0,18 der grundsätzlich positive Einfluss dieser Rubrik auf die Verweildauer zum Ausdruck. Besucher, die Seiten der Rubrik „Immobilien“ aufgerufen haben, nutzten das Portal im Mittel länger als Personen, die diese Seiten nicht besuchten. Die Inhalte dieser Rubrik spendeten den Besuchern offensichtlich einen merklichen Nutzen und verdienen es deshalb, auch weiterhin gepflegt zu werden.

of “eroticism”, “finances” and “music” gave reasons for longer dwell times, and thus obviously provided their visitors a comparatively high benefit. Figure 7: Normalized Parameter Estimators for the Internet Portal

The added value of the model-based analysis becomes evident, if one contrasts the parameter estimators presented in Figure 7 with the visit frequencies per category or subcategory given in Table 4. The category “real estate”, for example, shows only a comparatively small number of visits and hence represents one of the less frequently used contents of the portal. At the same time, the corresponding parameter estimator of 0.18 expresses the basic positive influence of this category on the dwell time. Visitors who selected pages of the category “real estate” – on average – used the portal longer than those who did not visit these pages. The contents of this category apparently provided the visitors with a significant benefit and therefore deserve to be maintained further on. The category “partner search” likewise shows only a small number of visits and, in addition, features a negative parameter value. The contents of the respective pages, at the time of data collection, seemingly have led to a reduction of the visiting times. This raises the question as to the causes of this effect. One possible reason could be the browsing of linked external Internet sites that did not belong to the portal and thus were not covered by the data. The identified effect could even have been of a positive nature in this case, for example, if the respective links had been banner advertisements for which the portal operator was compensated for per click. Another reason could be that visitors to the category “partner search” did not find the contents they expected there. Perhaps, some of the visitors expected erotic offers, which would suggest the need for improvement in terms of user guidance and/or communication of contents. The provided offers could of course also have been called (browsed) consciously, but did not correspond to the needs of the visitors in the end. In this case, a revision of the contents offered would be advisable.

Journal of Marketing Trends - Volume I (February 2011) 25


24366_int_Mise en page 1 22/12/10 12:25 Page26

JOURNAL OF MARKETING TRENDS - INTERNET

Die Rubrik „Partnersuche“ verfügt ebenfalls nur über eine geringe Besuchsfrequenz und zeichnet sich darüber hinaus durch einen negativen Parameterwert aus. Die Inhalte der betreffenden Seiten scheinen also zum Zeitpunkt der Datenerhebung eher zu einer Verkürzung der Besuchszeiten geführt zu haben. Damit stellt sich die Frage nach den Ursachen für diesen Effekt. Ein möglicher Grund könnte der Aufruf verlinkter externer Internetangebote sein, die nicht zum Portal gehörten und somit nicht erfasst wurden. Der identifizierte Effekt könnte in diesem Fall sogar positiver Natur gewesen sein, z. B. dann, wenn es sich bei den betreffenden Links um Bannerwerbung handelte und diese dem Portalbetreiber je Klick vergütet wurde. Ein anderer Grund könnte sein, dass Besucher der Rubrik „Partnersuche“ nicht die Inhalte vorfanden, die sie dort erwarteten. Möglicherweise vermuteten einige der Besucher hier auch erotische Angebote, was auf einen Verbesserungsbedarf hinsichtlich der Benutzerführung bzw. der Kommunikation der Inhalte hindeutet. Die bereitgestellten Angebote könnten aber natürlich auch bewusst aufgerufen worden sein, entsprachen dann jedoch nicht den Ansprüchen der Besucher. In diesem Fall wäre eine Überarbeitung der angebotenen Inhalte angeraten. Bei anderen Rubriken und Unterrubriken stellt sich die Sache gerade andersherum dar. Die Unterrubrik „Zeitvertreib“ beispielsweise weist eine vergleichsweise hohe Besuchszahl auf. Der Wert des betreffenden Parameters fällt hingegen eher klein aus, was darauf hindeutet, dass Besucher, die diese Unterrubrik aufsuchten, das Portal im Mittel schon nach kurzer Zeit wieder verließen. Dies ist vor allem deshalb bemerkenswert, weil davon auszugehen ist, dass für das Gros der Besucher, die dieses Angebot aufsuchten, die schnelle Suche und Aufnahme von Informationen wohl eher nicht im Vordergrund stand. Bereits diese wenigen Beispiele lassen erkennen, dass die modellbasierte Herangehensweise einen echten Mehrwert gegenüber der reinen Betrachtung von Besuchshäufigkeiten bieten kann. Der PoissonRegressionsansatz ermöglicht eine differenziertere Betrachtung, die explizit auf die Auswirkungen der vorgefundenen Inhalte in Form von Attraktionspotenzialen abzielt. Für eine umfassende Bewertung der angesprochenen Phänomene sind allerdings weiterführende Analysen erforderlich, die auch qualitative Aspekte mit einbeziehen. Der vorgeschlagene Ansatz hilft dabei, herauszufinden, welche Bereiche die Aufenthaltsdauer in einem Portal zukünftig erhöhen können und wo sich möglicherweise Verbesserungsbedarf abzeichnet. Eine Überarbeitung empfiehlt sich z. B. bei den Unterrubriken „Lifestyle/Gesundheit“ und „Partnersuche“. Möglicherweise sollte sogar in Erwägung gezogen werden, diese Unterrubriken ganz zu streichen,

26 ISSN 1961-7798 - © 2009, International Marketing Trends Conference

For other categories and subcategories, the situation presents itself the other way round. The subcategory “amusement”, for example, has a comparatively high number of visits. The estimate of the corresponding parameter, however, is rather small, which indicates that visitors to this subcategory already left the portal after a short time on average. This is especially noteworthy as it has to be assumed that for the bulk of visitors to these pages, a fast search and absorption of information rather did not stand in the foreground. Already these few examples show that model-based approaches can provide added-value to the mere consideration of visit frequencies. The Poisson regression approach allows for a more differentiated examination which explicitly focuses on the impact of the contents in terms of attraction potentials. However, for a comprehensive assessment of the mentioned phenomena, further analyses which also include qualitative aspects are needed. The suggested approach helps to identify those areas which could improve the dwell time in a portal in future and where a need for improvement emerges. So, for example, a revision of the subcategories “lifestyle/health” and “partner search” seems advisable. Maybe, it should even be considered to discard these subcategories and instead strive for a more directed focusing on the remaining offers. Such a reduction could, in particular, have a positive impact on the overall assessment of the portal, if it can be assumed that the visitors of the respective subcategories actually feel a negative benefit, for example, in terms of wasted time due to an emerging disorientation or insufficient contents. The results for the subcategory “weather” also deserve a closer look. The reading of a weather report, in contrast to the usage of more differentiated service offers, represents a utilization which is very limited in time. After the visitor has retrieved the weather data interesting to him, his information need is mostly satisfied. Thus, there is little reason for longer staying in this subcategory. Accordingly, the perceived utility of the portal as a whole can only be increased through the frequenting of other (sub-) categories. Consequently, it is important to firstly determine whether the visitors to the subcategory “weather” restrict themselves to the retrieval of weather information or whether they also use other service offers of the portal. In the first case, it turns out to be advantageous that the repeated and self-driven examination of an offer can cause a more positive perception of it. This phenomenon is discussed in the literature under the concept of “mere exposure effect” (Bornstein und Agostino 1992). Moreover, the insights gained from an application of the model can be beneficial regarding the customer retention potential of individual (sub-) categories, for example, when marketing corresponding advertising space.


24366_int_Mise en page 1 22/12/10 12:25 Page27

JOURNAL OF MARKETING TRENDS - INTERNET

und stattdessen eine stärkere Fokussierung des Angebots anzustreben. Eine solche Angebotsverschlankung könnte insbesondere dann positive Auswirkungen auf die Gesamtbeurteilung des Portals haben, wenn davon auszugehen ist, dass die Besucher der betreffenden Unterrubriken tatsächlich einen negativen Nutzen empfinden, z. B. im Sinne von Zeitverschwendung infolge einer entstandenen Desorientierung oder unzureichender Inhalte. Auch die Resultate zur Unterrubrik „Wetter“ verdienen eine genauere Betrachtung. Das Lesen eines Wetterberichts stellt im Gegensatz zur Inanspruchnahme stärker ausdifferenzierter Leistungsangebote eine Nutzung dar, die zeitlich stark begrenzt ist. Nachdem der Besucher die ihn interessierenden Wetterdaten abgerufen hat, ist sein Informationsbedarf zumeist befriedigt. Für einen längeren Aufenthalt in dieser Unterrubrik besteht somit kaum Veranlassung. Der wahrgenommene Nutzen des Portals als Ganzes kann somit erst durch das Aufsuchen anderer (Unter-) Rubriken erhöht werden. Folglich gilt es zunächst zu eruieren, ob sich die Besucher der Unterrubrik „Wetter“ tatsächlich zum überwiegenden Teil auf das Abrufen von Wetterinformationen beschränken oder ob sie auch weitere Leistungsangebote des Portals in Anspruch nehmen. Im ersten Fall erweist es sich als vorteilhaft, dass die wiederholte, aus eigenem Antrieb heraus erfolgende Auseinandersetzung mit einem Angebot eine positivere Wahrnehmung desselben hervorrufen kann. Dieses Phänomen wird in der Literatur unter dem Begriff des „mere exposure effect“ diskutiert (Bornstein und Agostino 1992). Darüber hinaus können die aus einer Modellanwendung resultierenden Erkenntnisse hinsichtlich des Kundenbindungspotenzials einzelner (Unter-) Rubriken von Nutzen sein, z. B. bei der Vermarktung entsprechender Werbeplätze.

4. Fazit und Ausblick Im vorliegenden Beitrag wurde ein flexibler PoissonRegressionsansatz zur Präferenzanalyse auf Basis von Nachfragedaten aus Internetplattformen vorgestellt. Dieses ermöglicht, unter Rückgriff auf bei der Inanspruchnahme von Internetangeboten anfallende Log-File-Daten, den Erhalt aussagekräftiger Informationen zur Attraktivität und Adäquanz bestehender Leistungsangebote. Die vorgeschlagene Methodik lässt Parallelen zur weit verbreiteten Conjoint-Analyse erkennen und ermöglicht für die Merkmalsausprägungen eines Leistungsangebotes (z. B. die verschiedenen Elemente einer Pauschalreise oder die einzelnen Unterrubriken eines Internetportals) die Schätzung von Parametern, die sich in einem weiteren Sinne als Teilnutzenwerte interpretieren lassen. Ihre maßnahmenorientierte Interpretation kann bei der Neu- oder Umgestaltung

4. Conclusion and Outlook In the present article, a flexible Poisson regression model for preference analysis based on demand data from Internet platforms is proposed. With the help of log file data as it accrues when using Internet offers, this approach enables the receipt of meaningful information about the attractiveness and adequacy of existing products or services. The methodology is similar to widely recognized conjoint analysis and enables the estimation of parameters for the attribute levels of a product or service offer (such as the different elements of a package tour or the individual subcategories of a web portal) which, in a broader sense, can be interpreted as part-worth utilities. Their action-oriented interpretation can be useful regarding the new development or modification of products or services, but without evoking additional costs for data collection. For vendors, the presented methodology offers a cost-effective possibility to obtain additional information on the preference structures of their customers. Thus, it can contribute to a deeper understanding of buying behaviour on the Internet. Furthermore, the model enables the identification of promising standard offers for those customers who evaluate product configuration systems negatively due to their complexity and therefore do not use them (Dellaert and Stremersch 2005). Of course, the suggested approach also has its shortcomings: The model can only be used to determine preferences for those attribute levels which are already available on the market. In addition, the approach is based on data from single online service providers. Consequently, the preferences measured in this way can be different from those of the users of other providers. A promising extension of the presented approach is the explicit consideration of the heterogeneity of individual usage behaviour by means of a suitable distribution for λ (e.g., a negative binomial distribution). Such approaches already have a long tradition in stochastic modelling of buying behaviour (Decker and Wagner 2002). Through such a model extension, the problem of heterogeneity of individual preferences could be taken into account in an elegant way.

Journal of Marketing Trends - Volume I (February 2011) 27


24366_int_Mise en page 1 22/12/10 12:25 Page28

JOURNAL OF MARKETING TRENDS - INTERNET

von Produkten und Services von Nutzen sein, ohne im Gegenzug zusätzliche Erhebungskosten hervorzurufen. Für Unternehmen bietet die vorgestellte Methodik eine kostengünstige Möglichkeit, um zusätzliche Informationen über die Präferenzstrukturen ihrer Kunden zu erhalten. Sie kann damit zu einem tieferen Verständnis des Kaufverhaltens im Internet beitragen. Darüber hinaus ermöglicht das Modell die Identifizierung Erfolg versprechender Standardangebote für jene Kunden, die Produktkonfiguratoren aufgrund ihrer Komplexität negativ bewerten und deshalb nicht verwenden (Dellaert und Stremersch 2005). Natürlich ist auch der vorgeschlagene Ansatz nicht frei von Schwächen: So können damit nur für jene Leistungsausprägungen Präferenzen bestimmt werden, die bereits am Markt verfügbar sind. Des Weiteren basiert der Ansatz auf Daten einzelner Online-Anbieter. Die auf diese Weise gemessenen Präferenzen können sich folglich von denen der Nutzerschaft anderer Anbieter unterscheiden. Eine vielversprechende Erweiterung des vorgestellten Ansatzes stellt die explizite Berücksichtigung der Heterogenität des individuellen Nutzungsverhaltens mittels einer geeigneten Verteilung für dar (z. B. einer Negativ-Binomial-Verteilung). Entsprechende Vorgehensweisen haben in der stochastischen Kaufverhaltensmodellierung bereits eine lange Tradition (Decker und Wagner 2002). Durch eine solche Modellerweiterung könnte dem Problem der Heterogenität individueller Präferenzen auf elegante Weise Rechnung getragen werden.

Literatur / References Backhaus, K., T. Hillig und R. Wilken (2007), “Predicting Purchase Decisions with Different Conjoint Analysis Methods: A Monte Carlo Simulation,” International Journal of Market Research, Jg. 49, Nr. 3, 341-364. Baier, D. und M. Brusch (2009): Conjointanalyse – Methoden – Anwendungen – Praxisbeispiele, Berlin: Springer. Blossfeld, H.-P. und G. Rohwer (1995), Techniques of Event History Modeling. Mahwah: Lawrence Erlbaum Associates. Bornstein, R. F. und P. R. D’Agostino (1992), “Stimulus Recognition and the Mere Exposure Effect,” Journal of Personality and Social Psychology, Jg. 63, Nr. 4, 545-552. Burklen, S., P. J. Marron, S. Fritsch und K. Rothermel (2005), “User Centric Walk: An Integrated Approach for Modeling the Browsing Behavior of Users on the Web,” in: Proceedings of the 38th Annual Symposium on Simulation, IEEE Computer Society, 149-159.

28 ISSN 1961-7798 - © 2009, International Marketing Trends Conference

Cheung, C. M. K., L. Zhu, T. Kwong, G. W. W. Chan und M. Limayem (2003), “Online Consumer Behavior: A Review and Agenda for Future Research,” in: Proceedings of the 16th Bled eCommerce Conference, Bled, 194-218. Constantinides, E. (2004), “Influencing the Online Consumer’s Behavior: The Web Experience,” Internet Research, Jg. 14, Nr. 2, 111-126. Cooley, R., B. Mobasher und J. Srivastava (1999), “Data Preparation for Mining World Wide Web Browsing Patterns,” Journal of Knowledge and Information Systems, Jg. 1, Nr.1, 5-32. Dahan, E. und V. S. Srinivasan (2000), “The Predictive Power of Internet-Based Product Concept Testing Using Visual Depiction and Animation,” Journal of Product Innovation Management, Jg. 17, Nr. 3, 99109. Decker, R. und R. Wagner (2002), Marketingforschung – Methoden und Modelle zur Bestimmung des Käuferverhaltens. München: Moderne Industrie. Decker, R. und S. W. Scholz (2008), „Modellgestützte Nachfrageanalyse im Internet als Grundlage für eine kundenorientierte Angebotspolitik,“ in: Proceedings of the 7th International Congress „Marketing Trends“, Venice. Dellaert, B. G. C. und S. Stremersch (2005), “Marketing Mass-Customized Products: Striking a Balance between Utility and Complexity,” Journal of Marketing Research, Jg. 42, Nr. 2, 219-227. Drèze, X. und F. Zufryden (1997), “Testing Web Site Design and Promotional Content,” Journal of Advertising, März/April, 77-91. Drèze, X. und F. Zufryden (1998), “A Web-based Methodology for Product Design Evaluation and Optimisation,” Journal of the Operational Research Society, Jg. 49, September, 1034-1043. Kuhfeld, W. F. (2003), Marketing Research Methods in SAS: Experimental Design, Choice, Conjoint, and Graphical Techniques. Cary NC: SAS Institute Inc. Kuom, M. und B. Oertel (1999), “Virtual Travel Agencies,” Netnomics, Jg. 1, 225-235. Law, R., K. Leung und J. Wong (2004), “The Impact of the Internet on Travel Agencies,” International Journal of Contemporary Hospitality Management, Jg. 16, Nr. 2, 100-107. Long, J. S. und J. Freese (2006), Regression Models for Categorical Dependent Variables Using Stata. 2. Aufl., College Station: Stata Press.


24366_int_Mise en page 1 22/12/10 12:25 Page29

JOURNAL OF MARKETING TRENDS - INTERNET

Meißner, M., S. W. Scholz und R. Decker (2007), “Analytic Hierarchy Process vs. Adaptive Conjoint Analysis – An Empirical Comparison,” in: Data Analysis, Machine Learning, and Applications, Preisach, C. et al., Hrsg., Heidelberg: Springer, 447-454. Meißner, M., S. W. Scholz und R. Wagner (2008), “Marketing Research Using Multimedia Technologies,” Encyclopedia of Multimedia Technology and Networking, Pagani, M., Hrsg., Hershey: Idea Group Publishing, 880-886. Montgomery, A. L., S. Li, K. Srinivasan und J. C. Liechty (2004), “Modeling Online Browsing and Path Analysis Using Clickstream Data,” Marketing Science, Jg. 23, Nr. 4, 579-595. Oertel, B., T. Feil und S. L. Thio (2003), „Neue IuKTechnologien und ihre Relevanz für die Wettbewerbsfähigkeit touristischer Destinationen,“ Bericht Nr. 56, Institut für Zukunftsstudien und Technologiebewertung, Berlin. Scholz, S. W., M. Meißner und R. Decker (2009): “Measuring Consumer Preferences for Complex Products: A Compositional Approach Based on Paired Comparisons,” erscheint in Journal of Marketing Research. Spiliopoulou, M. (2000), “Web Usage Mining for Web Site Evaluation,” Communications of the ACM, Jg. 43, Nr. 8, 127-134. Steckel, J. H., R. S. Winer, R. E. Bucklin, B. G. C. Dellaert, X. Drèze, G. Häubl, S. D. Jap, J. D. Little, T. Meyvis, A. L. Montgomery und A. Rangaswamy (2005), “Choice in Interactive Environments,” Marketing Letters, Jg. 16, Nr. 3-4, 309-320. Tao, Y.-H., T.-P. Hong und Y.-M. Su (2006), “Improving Browsing Time Estimation with Intentional Browsing Data,” International Journal of Computer Science and Network Security, Jg. 6, Nr. 12, 35-39. Venkatesh, V., R. Agarwal (2006), “Turning Visitors into Customers: A Usability-Centric Perspective on Purchase Behavior in Electronic Channels,” Management Science, Jg. 52, Nr. 3, 367-382.

Journal of Marketing Trends - Volume I (February 2011) 29


24366_int_Mise en page 1 22/12/10 12:25 Page30

30 ISSN 1961-7798 - Š 2009, International Marketing Trends Conference


24366_int_Mise en page 1 22/12/10 12:25 Page31

J O U R N A L O F M A R K E T I N G T R E N D S - L O YA LT Y P R O G R A M

Programme de fidélité associant deux détaillants concurrents : effets sur le comportement d’achat des utilisateurs Loyalty program bringing together two competing retailers: effects on users’ shopping behavior

Le but de cet article est d’étudier les comportements des souscripteurs à un programme de fidélité envers deux enseignes concurrentes participantes. La méthodologie utilisée repose sur un modèle d’équations structurelles de courbes latentes. Les trajectoires d’achat des utilisateurs auprès des deux enseignes concurrentes ne sont pas significativement différentes. Aussi bien trajectoires d’achat des clients exclusifs de chaque enseigne que trajectoires d’achat des clients multi-fidèles. Les tendances des clients multi-fidèles, à acheter auprès de chacune des deux enseignes sont indépendantes l’une de l’autre. The aim of this paper is to study the behavior of subscribers to a loyalty program in which two of the participants are stores of competing retail chains. The methodology used is latent curve model using structural equations. The users’ purchasing trajectories at the two competing stores are not significantly different. This is true both for customers who shopped exclusively in one or other store and for customers who used both. The tendencies of multi-loyal customers to shop at each of the two stores are independent of each other. Mots clés : programme de fidélité multi-sponsors, tendance latente à l’achat initial, tendance latente au contrôle de l’achat initial. Key words: multi-sponsor loyalty program, latent initial purchase tendency, latent tendency to control the initial purchase.

> Jean Frisou INSEEC Bordeaux jfrisou@groupeinseec.com

> Hélène Yildiz Université de Nancy II helene.yildiz@univ-nancy2.fr

> Laurent Flores INSEEC Paris lflores@groupeinseec.com Journal of Marketing Trends - Volume I (February 2011) 31


24366_int_Mise en page 1 22/12/10 12:25 Page32

J O U R N A L O F M A R K E T I N G T R E N D S - L O YA LT Y P R O G R A M

Introduction

Introduction

Les programmes de fidélité sont devenus d’un usage courant dans la plupart des activités commerciales. Si leur efficacité est loin de faire l’unanimité parmi les chercheurs, les praticiens ne veulent retenir de leurs travaux que les avancées positives. Ils évoquent les effets des récompenses sur la rétention des clients (Verhoef 2003), sur la part des dépenses qu’ils consacrent aux sponsors (Verhoef 2003 ; Meyer-Waarden 2007) ou encore sur la durée de leurs relations avec ces sponsors (Meyer-Waarden 2007).

Loyalty programs have become commonly used in most commercial businesses. While their effectiveness is far from generally accepted among researchers, practitioners are inclined to adopt only the positive conclusions from these studies. They point to the effects of rewards on retaining customers (Verhoef 2003), share-of-wallet devoted to sponsors (Verhoef 2003; Meyer-Waarden 2007) and the duration of their relationships with these sponsors (Meyer-Waarden 2007).

De nombreux auteurs situent les sources d’efficacité de ces programmes dans les systèmes de récompenses qui les supportent (Dowling et Uncles 1997; Roehm, Pullins and Roehm Jr 2002, Yi et Jeon 2003 ; Leenheer, van Heerde, Bijmolt and Smidts 2007). L’idée de bâtir des coalitions de sponsors pour offrir aux clients, dans des délais plus rapides, des récompenses plus attractives et plus diversifiées s’est très vite imposée (Reinares, Madariaga 2007). Le programme Nectar de Loyalty Management UK conçu en 2002 réunit ainsi onze partenaires. Toutefois ces programmes n’associent le plus souvent que des enseignes dont les offres se complètent et qui ne se font jamais directement concurrence. Le programme Nectar par exemple est adossé à des enseignes très diverses : Sainsbury’s, Barclaycard, Debenhams, BP, Threshers, Vodafone, Adams, e-Energy, Allsports, Winemark and Ford. Seul, S’Miles à notre connaissance réunit des enseignes concurrentes, distribuant les mêmes catégories de produits (Monoprix et Géant Casino). Mais leur concurrence n’est pas vraiment directe car Monoprix s’implante au cœur des villes les plus importantes, alors que Géant Casino s’installe à la périphérie des villes de moyenne importance. Pour étudier l’efficacité des programmes de fidélité, les chercheurs ont jusqu’à présent considéré des programmes propriétaires et des programmes multisponsors qui excluent des enseignes concurrentes. Mais comment évolueraient les comportements des utilisateurs d’un programme de fidélité mutualisé si celui-ci regroupait en un même lieu, deux hypermarchés directement concurrents tels qu’Auchan et Carrefour ? La question peut paraître singulière, mais c’est ainsi qu’elle se pose aux animateurss des programmes de fidélité, conçus pour revitaliser les centres villes. Faut il intégrer dans un même programme de fidélité plusieurs pâtisseries concurrentes présentes dans le même centre ville ? Les clients fidèles à chacun de ces magasins ne seraient ils pas alors encouragés par le programme à rendre visite à leurs concurrents ? Le programme ne protègerait peut être plus aussi efficacement les sponsors partenaires mais néanmoins concurrents. Nous sommes dès lors fondé à poser la question de recherche suivante :

32 ISSN 1961-7798 - © 2009, International Marketing Trends Conference

Many authors locate the sources of the effectiveness of these programs in the reward systems supporting them (Dowling and Uncles 1997; Roehm, Pullins and Roehm Jr 2002, Yi and Jeon 2003; Leenheer, van Heerde, Bijmolt and Smidts 2007). The idea of building sponsor coalitions to offer customers more attractive and varied rewards within as short a time period as possible caught on very quickly (Reinares, Madariaga 2007). The Loyalty Management UK Nectar program, created in 2002, thus brings together eleven partners. However, these programs mostly frequently involve companies whose offerings complement each other and are not directly in competition. The Nectar program, for example, is backed by very diverse companies: Sainsbury’s, Barclaycard, Debenhams, BP, Threshers, Vodafone, Adams, e-Energy, Allsports, Winemark and Ford. To our knowledge, only S’Miles includes companies – Monoprix and Géant Casino – who retail the same product categories. But their competition is not really direct, since Monoprix outlets are located in the center of large cities, while Géant Casino outlets are found on the outskirts of medium-sized towns. To study the effectiveness of loyalty programs, until now researchers have considered proprietary and multi-sponsor programs that exclude competing companies. But how would behaviors of a shared loyalty program evolve if this involved two directly competing hypermarkets in the same place, such as Auchan and Carrefour ? The question may appear to be strange, but this is precisely what developers of loyalty programs designed to revitalize city centers are wondering. Should the same loyalty program include several competing cake shops in the same city ? Might not customers loyal to a particular shop be encouraged by the program to visit its competitors ? Perhaps the program would no longer as effectively protect the competing partner sponsors. Thus we are justified in raising the following research question. What are the effects of multi-sponsor loyalty programs on users’ purchasing behavior, when the programs include two retailers which are in direct competition? This paper aims to answer this question by not only putting forward theoretical proposals but also by providing


24366_int_Mise en page 1 22/12/10 12:25 Page33

J O U R N A L O F M A R K E T I N G T R E N D S - L O YA LT Y P R O G R A M

Quels sont les effets des programmes de fidélité multisponsors sur les comportements d’achat des utilisateurs, lorsque ces programmes associent deux enseignes en concurrence directe ? L’objectif poursuivi dans cet article est de répondre à cette question, en apportant des propositions théoriques mais aussi des confirmations empiriques. Nous défendons l’idée selon laquelle, la présence de concurrents dans un même programme de fidélité multienseignes, n’induit pas de différences significatives dans le comportement d’achat des clients auprès de chacun d’eux. Dans une première partie nous reviendrons sur les concepts et les théories de la fidélité à partir desquels nous organisons notre réflexion. Dans une seconde partie nous justifierons et nous préciserons nos hypothèses de recherche. Une troisième partie intitulée « méthodologie » nous permettra de les tester. Enfin nous conclurons dans une dernière partie en dégageant, les apports, les limites et les perspectives managériales de ce travail.

CADRE CONCEPTUEL Investiguer l’efficacité des programmes de fidélité nécessite de remettre en perspective leurs fondements conceptuels et théoriques. Nous aborderons les deux principaux thèmes sur lesquels la littérature s’est polarisée, fidélité puis fidélisation, en soulignant l’opposition des points de vues qu’ils recèlent. A partir des propositions les plus récentes formulées dans la littérature, nous esquisserons une définition dynamique du concept de fidélité, propre à rendre compte de l’évolution des comportements d’achats des clients utilisant un programme.

La fidélité du consommateur La fidélité du consommateur est un concept marketing marqué par des oppositions très fortes entre deux courants de recherche, l’un empiriste, l’autre essentialiste. Le premier définit la fidélité comme le résultat de l’expérience sensible que sont les achats répétés (Tucker 1964; Stafford 1966; McConnel 1968 ; Ehrenberg 2000). La théorie béhavioriste de l’apprentissage instrumental ou opérant explique alors le phénomène (Filser 1994). Le consommateur tend à reproduire les achats qui lui procurent une satisfaction et tend à éviter les achats qui lui procurent une déception, Les achats sont donc renforcés les uns par les autres du fait de leurs conséquences. Le second courant affirme au contraire que l’essence de la fidélité précède et ses expériences. Pour justifier leur approche les auteurs font observer que le comportement d'achat répété n'est qu'une condition nécessaire mais non suffisante de la fidélité. (Day, 1969), Jacoby et Kyner (1973) suggèrent donc une définition de la fidélité, à partir de ses causes et de ses conséquences. La fidélité est: 1) une réponse comportementale (i,e,: un achat), 2) biaisée (i,e,: non aléatoire), 3) exprimée au cours

empirical confirmation. We defend the idea that the presence of competitors in the same multi-sponsor loyalty program does not result in significant differences in customers’ purchasing behavior at each competitor. In the first part of the paper we shall revisit the concepts and theories of loyalty on which we base our thinking. In the second part we shall justify and specify our research hypotheses. The third part, Methodology, will enable us to test these. Finally, we will conclude in last part by drawing out the contributions, limitations and managerial perspectives of this study.

CONCEPTUAL FRAMEWORK Investigating the effectiveness of loyalty programs requires putting their conceptual and theoretical bases into perspective. We shall first deal with the two main themes around which the literature is polarized, namely loyalty and loyalty creation, by emphasizing the opposing viewpoints they embody. From the most recent suggestions in the literature, we shall outline a dynamic definition of loyalty, suitable for taking account of the evolution of the purchasing behavior of customers using a program.

Consumer loyalty Consumer loyalty is a marketing concept that is characterized by very pronounced oppositions between two research approaches, one of them empiricist, the other essentialist. The former defines loyalty as the result of the lived experience of making repeat purchases (Tucker 1964; Stafford 1966; McConnel 1968; Ehrenberg 2000). The behaviorist theory of instrumental or effective learning then explains the phenomenon (Filser 1994). Consumers tend to repeat purchases that bring them satisfaction and tend to avoid purchases that result in disappointment. Successive purchases are therefore reinforced through their consequences. The second approach, in contrast, argues that the essence of loyalty precedes consumers’ purchasing experiences. To justify this approach its defenders point out that repeated purchasing is merely a necessary but not a sufficient condition for loyalty (Day, 1969). Jacoby and Kyner (1973) therefore put forward a definition of loyalty based on its causes and its consequences. Loyalty is: 1) a behavioral response (i.e. a purchase), that is 2) biased (i.e. nonrandom), 3) and expressed over time, 4) by a decision unit, 5) in regard to one or more alternatives appearing in a set of brands, 6) and which is the result of psychological processes (i.e. decision-making, evaluation). Other authors say the same thing, but much more simply: “a consumer is loyal to a brand if he buys it regularly and has developed a favorable attitude toward it” (Filser 1994) or, again, “loyalty is expressed by consumption behaviors and is explained by consumers’ favorable attitudes toward the brand’s products” (Trinquecoste 1996).

Journal of Marketing Trends - Volume I (February 2011) 33


24366_int_Mise en page 1 22/12/10 12:25 Page34

J O U R N A L O F M A R K E T I N G T R E N D S - L O YA LT Y P R O G R A M

du temps, 4) par une unité de décision, 5) au regard d'une ou plusieurs alternatives figurant dans un ensemble de marques, 6) et qui est le résultat de processus psychologiques (i,e,: prise de décision, évaluation). D’autres auteurs diront la même chose, mais beaucoup plus simplement : « un consommateur est fidèle à une marque s’il l’achète régulièrement et qu’il a développé à son égard une attitude favorable » (Filser 1994) ou encore « la fidélité s’exprime par les comportements de consommation et s’explique par les attitudes favorables des consommateurs à l’égard des produits de la marque » (Trinquecoste 1996). Sans une motivation forte donnant à l'achat répété le sens d’un achat désiré, la répétition d’achat est considérée comme une simple inertie, qui n’est autre qu’une fidélité fallacieuse (Day, 1969). Pour caractériser la fidélité "vraie", les chercheurs ont proposé de lui associer des marqueurs psychologiques. Day (1969) a suggéré l'attitude envers la marque. Jacoby et Kyner (1973) ont avancé l'engagement qui révèle deux facettes distinctes. L'engagement affectif et l'engagement cognitif ou calculé (Amine 1998). Mais l'attitude et l'engagement ne sont tournés que vers la marque dont ils sont supposés expliquer la fidélité, alors que le processus d'évaluation évoqué par Jacoby et Kyner (1973) suppose des comparaisons entre plusieurs alternatives. Pour résoudre cette difficulté, Dick et Basu (1994) ont suggéré de considérer l'attitude relative envers la marque plutôt que l’attitude absolue. Odin (1998) a proposé de recourir au concept plus général de sensibilité du consommateur à la marque. Un client est sensible à la marque, si le fait qu’un produit soit une marque ou pas, le fait changer de comportement d’achat (Kapferer et Laurent 1983).

Programmes de fidélité La définition des programmes de fidélité, n’échappe pas à l’opposition entre approches béhavioriste et cognitiviste de la fidélité. Les programmes de fidélité ne sont d’ailleurs pas le produit d’une réflexion académique, bien qu’ils s’inspirent dans leur principe de la théorie béhavioriste de l’apprentissage opérant (Foxall 1996, 1997). Deux perspectives béhavioriste et cognitiviste suggèrent des définitions différentes d’un programme de fidélité. La perspective béhavioriste calque ses définitions sur la manière dont sont réalisés ces programmes. Elles mettent l’accent sur les incitations et les récompenses qui sont offertes au consommateur en contrepartie de ses achats répétés. Un programme de fidélité est défini comme un programme marketing conçu pour construire la fidélité du client en lui fournissant des incitations et des récompenses (Yi et Jeon 2003). Cette définition rejoint celle de Liu (2007), pour qui, un programme de fidélité est un programme qui permet aux consommateurs d’accumuler des récompenses gratuites quand il a un

34 ISSN 1961-7798 - © 2009, International Marketing Trends Conference

Without a strong motivation bringing a sense of active desire to the repeat purchase, such repetition is viewed as simple inertia, or false loyalty (Day, 1969). To characterize loyalty as “true” loyalty, researchers have proposed linking it to psychological markers. Day (1969) suggested attitude toward the brand. Jacoby and Kyner (1973) suggested commitment, which reveals two distinct aspects: affective commitment and cognitive or calculated commitment (Amine 1998). But attitude and commitment are directed only at the brand the loyalty to which they are supposed to explain, whereas the evaluation process referred to by Jacoby and Kyner (1973) assumes comparisons of several alternatives. To resolve this problem, Dick and Basu (1994) proposed considering relative rather than absolute attitude toward the brand. Odin (1998) suggested using the more general concept of consumer sensitivity to the brand. A customer is brand sensitive if the fact that a product is a brand or not a brand makes him change his purchasing behavior (Kapferer and Laurent 1983).

Loyalty programs The definition of loyalty programs also involves the opposition between behaviorist and cognitivist approaches to loyalty. Loyalty programs are in any case not the product of academic reflection, although their principle is inspired by the behaviorist theory of effective learning (Foxall 1996, 1997). Both behaviorist and cognitivist approaches give rise to their own characteristic definitions of a loyalty program. The behaviorist perspective takes its definitions from the way in which they are implemented in these programs. These place the emphasis on the incentives and rewards that are offered to consumers in compensation for repeated purchases. A loyalty program is defined as a marketing program designed to build customer loyalty by providing customers with incentives and rewards (Yi and Jeon 2003). This definition is similar to that given by Lui (2007), for whom a loyalty program is a program that enables consumers to accumulate free rewards when a purchasing behavior is repeated. We note the highly descriptive character of these definition, which, moreover, reveal certain weaknesses. On the one hand they reduce loyalty programs to reward programs, whereas loyalty could be obtained by other means. On the other hand the dependence linkage between repeat purchases and rewards suggests that the effect of a loyalty program is automatic. If repeat purchases lead to rewards, the rewards obtained will not automatically result in a continuation of repeat purchasing. The literature thus abounds in examples showing that even if the effects of loyalty programs are positive, they are not very significant. The pioneering study by Sharp and Sharp (1997) reveals a weak impact of loyalty


24366_int_Mise en page 1 22/12/10 12:25 Page35

J O U R N A L O F M A R K E T I N G T R E N D S - L O YA LT Y P R O G R A M

comportement d’achat répété. On notera le caractère très descriptif de ces définitions, qui présentent par ailleurs plusieurs défauts. D’une part elles réduisent les programmes de fidélité à des programmes de récompenses, alors que la fidélité pourraient être obtenue par d’autres voies. D’autre part le lien de dépendance entre achat répété et récompenses peut laisser croire que l’effet d’un programme de fidélité est automatique. Si l’achat répété induit des récompenses, les récompenses obtenues n’induiront pas automatiquement le maintien de l’achat répété. La littérature regorge ainsi d’exemples montrant que même quand ils sont positifs les effets des programmes de fidélité ne sont pas très importants. L’étude pionnière de Sharp et Sharp (1997) révèle un faible impact des programmes de fidélité sur le comportement d’achat. En outre seulement une faible proportion des clients qui adhérent à un programme de fidélité sont vraiment fidèles à ce programme (Mauri 2003 , Allaway, Gooner, Berkowitz, et Davis (2006). De nombreux chercheurs ont d’ailleurs exprimé leur doute au sujet de l’efficacité des programmes de récompenses (O’Brien et Jones 1995 ; Dowling et Uncles 1997 ; Keh et Lee 2006 ; Leenheer, van Heerde , Bijmolt, Smidts 2007). Shugan (2005) dénonce à ce sujet l’amalgame qui est fait entre programmes de récompenses et programmes de fidélité. Il suggère sur le plan conceptuel, une redéfinition des programmes de fidélité. L’auteur rappelle que la motivation du client pour la marque ou pour l’entreprise est l’élément fondateur d’une relation durable conduisant à la fidélité. Cette motivation ou prédisposition en faveur de la marque transforme le client en un actif pour l’entreprise. Or, remarque l’auteur, les programmes de fidélité tels qu’ils existent aujourd’hui promettent aux clients des récompenses futures en contrepartie de leurs achats courants. Ils font d’eux non pas des actifs mais des dettes futures pour l’entreprise. Plutôt que de témoigner au client sa confiance en lui manifestant un engagement sans faille, l’entreprise demande au client de lui faire a priori confiance. L’auteur suggère donc une redéfinition des programmes de fidélité faisant une plus large place aux aspects cognitifs de la fidélité. Un « vrai » programme de fidélité ne doit pas se réduire selon Shugan à un système de règles associant des achats courants répétés à des bénéfices futurs. Un « vrai » programme de fidélité doit selon lui créer un actif, en rendant au cours du temps le client plus attaché à la marque. Cet attachement soutient l’auteur, peut provenir de l’apprentissage, de l’expérience, de la familiarité, de l’habitude, d’un coût de changement plus élevé ou de la personnalisation du programme.

MODÈLE DE RECHERCHE

programs on purchasing behavior. In addition, only a small proportion of customers who belong to a loyalty program are genuinely loyal to this program (Mauri 2003, Allaway, Gooner, Berkowitz and Davis (2006). Many researchers have also expressed their doubts on the subject of the effectiveness of loyalty programs (O’Brien and Jones 1995; Dowling and Uncles 1997; Keh and Lee 2006; Leenheer, van Heerde, Bijmolt, Smidts 2007). On this subject Shugan (2005) criticizes the confusion between reward programs and loyalty programs. He suggests that at a conceptual level loyalty programs should be redefined. He points out that the customer’s motivation for the brand or for the company is the founding element for a lasting relationship leading to loyalty. Such motivation or predisposition in the favor of the brand transforms the customer into an asset for the company. Yet loyalty programs such as they exist today, Shugan says, promise customers future rewards in return for their current purchases. This makes them not assets but future debts for the company. Rather than testifying to its trust in customers by showing its unfailing commitment to them, the company asks customers to place their trust in it from the outset. Shugan therefore suggests redefining loyalty programs by placing a greater emphasis on the cognitive aspects of loyalty. A “genuine” loyalty program should not, according to Shugan, reduce to a system of rules linking current purchases to future benefits. Rather it should create an asset, by making the customer more attached to the brand over time. This attachment, Shugan maintains, can come from learning, experience, familiarity, habit, a higher cost of switching brands, or the personalization of the program.

RESEARCH MODEL Contributions from the literature The opposition between genuine loyalty and false loyalty (Day 1969; Jacoby and Kyner 1973; Dick and Basu 1994; Amine 1998 ), and a genuine loyalty program and an alleged loyalty program (Shugan 2005), adds nothing to an understanding of the psychological processes at work. Indeed, ever since Day (1969) we know that loyalty has internal causes and external manifestations. Nevertheless there are many researchers who call for a development of modes of conceptualization of loyalty behavior, so that they can take better account of its temporal dimension (Kopalle and Neslin 2003; Taylor and Neslin 2005; MeyerWarden 2007; Liu 2007). We therefore propose a dynamic conceptual framework for loyalty that incorporates its observable behaviors and its latent cases. To explain the genesis of this theoretical framework, we draw on Figure 1, which shows how it complements the behaviorist and attitudinal paradigms of loyalty that preceded it.

Les apports de la littérature Les oppositions entre fidélité vraie et fidélité fallacieuse (Day 1969 ; Jacoby et Kyner 1973 ; Dick et Basu 1994 ;

From the behaviorist loyalty paradigm we retain the repeated purchasing behaviors in which loyalty is Journal of Marketing Trends - Volume I (February 2011) 35


24366_int_Mise en page 1 22/12/10 12:25 Page36

J O U R N A L O F M A R K E T I N G T R E N D S - L O YA LT Y P R O G R A M

Amine 1998 ), de même qu’entre programme de fidélité vrai et prétendu programme de fidélité (Shugan 2005), n’apportent rien à la compréhension des processus psychologiques qui sont à l’œuvre. On sait en effet depuis Day (1969) que la fidélité a des causes internes et des manifestations externes. Nombreux toutefois sont les chercheurs qui appellent à une évolution des modes de conceptualisation du comportement de fidélité, pour qu’ils intègrent davantage la dimension temporelle du phénomène (Kopalle et Neslin 2003 ; Taylor et Neslin 2005 ; MeyerWarden 2007 ; Liu 2007). Nous proposons donc un cadre conceptuel dynamique de la fidélité qui incorpore ses comportements observables et ses causes latentes . Pour expliquer la genèse de ce cadre théorique, nous nous appuyons sur la figure 1 qui montre comment il complète les paradigmes béhavioriste et attitudinal de la fidélité qui l’ont précédé.

reflected (Figure 1A). However, this is not a matter of a few dozen purchases, but of purchases made over a period of several years. Figure 1. A new conceptual framework for conceptualizing loyalty

Nous conservons du paradigme béhavioriste de la fidélité, les comportements d’achats répétés dans lesquels la fidélité se reflète (Figure 1 A). Toutefois il ne s’agit pas de quelques dizaines d’achats, mais des achats constatés sur une longue période, durant plusieurs années. Figure 1. A new conceptual framework for conceptualizing loyalty

From the cognitive paradigm we take the idea according to which buying behaviors are partly under the internal control of the individual. But instead of situating this control in the attitude or the commitment to the brand, we deploy the concept of latent tendency. Indeed Dall' Olmo Riley et al. (1997) showed that attitudinal variables were very unstable over time. It therefore seems inappropriate to adopt it for indicating the effect of loyalty of repeat purchase behaviors in the long term.

Nous empruntons au paradigme cognitif l’idée selon laquelle ces comportements d’achat sont en partie placés sous le contrôle interne de l’individu. Mais au lieu de situer ce contrôle dans l’attitude ou dans l’engagement avec la marque nous mobilisons le concept de tendance latente, Dall' Olmo Riley et alii (1997) ont en effet montré que les variables attitudinales étaient très instables au cours du

36 ISSN 1961-7798 - © 2009, International Marketing Trends Conference

The concept of latent tendency seems to us to be more appropriate. This is defined in psychology as “response probability” that very likely but not certainly will produce the expected behavior (Coutu 1949; Skinner 1965). In our theoretical model of tendential loyalty, the consumer’s internal control makes three tendencies occur (Figure 1C): first, a latent tendency to initial purchasing behavior that contributes to establishing the use level sought; second, a latent tendency by the consumer to control the evolution of the initial purchase level in subsequent periods. If the tendency to control is positive or zero, i.e. if the initial purchase level increases or remains unchanged over time, we shall speak of a loyalty behavior tendency. If neither of these is the case, then the tendency is toward disloyal behavior.


24366_int_Mise en page 1 22/12/10 12:25 Page37

J O U R N A L O F M A R K E T I N G T R E N D S - L O YA LT Y P R O G R A M

temps. Il semble donc inopportun d’y recourir pour marquer l’empreinte de la fidélité sur les comportements d’achat répétés sur le long terme. Le concept de tendance latente nous paraît plus approprié. Il est défini en psychologie comme une «probabilité de réponse» qui produira vraisemblablement mais pas certainement le comportement attendu (Coutu 1949 ; Skinner 1965). Dans notre modèle théorique de la fidélité tendancielle, le contrôle interne du consommateur fait intervenir trois tendances (Figure 1C). Une tendance latente au comportement d’achat initial qui contribue à fixer le niveau d’usage recherché. Une tendance latente du consommateur à contrôler dans les périodes suivantes l’évolution du niveau initial d’achat. Si la tendance au contrôle est positive ou nulle, c’est à dire si le niveau d’achat initial est augmenté ou se maintient au cours du temps nous parlerons de tendance au comportement de fidélité. Dans le cas contraire de tendance au comportement d’infidélité. Par exemple, lorsqu’un consommateur adhère à un programme de fidélité il apprend durant les six mois les règles du programme et détermine son niveau initial d’achat. Les mois suivants il progressera ou non dans son apprentissage, maîtrisera ou non de mieux en mieux les règles d’obtention des points, et en définitive augmentera ou diminuera le niveau initial d’achat. Un autre facteur déterminant du comportement d’achat est la tendance plus ou moins prononcée du client à composer avec la situation d’achat. Ce point particulier est mis en exergue par plusieurs auteurs. Les chercheurs considèrent souvent à tort que l’achat est réalisé dès lors que l’individu l’a décidé. Bagozzi et Warshaw (1990). proposent au contraire de considérer les comportements d’achat comme des essais ou comme des tentatives, qui parfois se heurtent à des empêchements, liés à la situation d’achat. Par exemple, les ruptures de stock en magasin ou l’impossibilité de recourir à un crédit peuvent contrarier les intentions du consommateur. Oliver (1999) insiste lui aussi sur la nécessité de prendre en compte ces obstacles quand on envisage de définir la fidélité. Pour Oliver (1999) le consommateur fidèle est celui qui contrôle ces obstacles et qui tente de les surmonter. Nous reprenons cette idée dans notre cadre théorique en introduisant lors de chaque situation la tendance du consommateur à contrôler les situations qui se présentent à lui. En résumé, le cadre éclectique que nous proposons coordonne des théories que les auteurs ont trop souvent tendance à opposer. Il intègre aussi des propositions d’amélioration du concept de fidélité suggérées dans la littérature (Figure 1C). Comme Jacoby et Kyner (1971) nous y invitent, nous définissons la fidélité à la fois par ses conséquences, que sont les comportements observés, et par ses causes, qui poussent le client à contrôler l’évolution de son niveau initial d’achat. Mais

For example, when consumers join a loyalty program they learn the rules of the program during the first six months and decide on their initial purchasing level. In the following months they will progress (or not) in this learning process, will better master the rules for obtaining points (or fail to do so), and in fact will increase or reduce their initial purchasing level. Another factor determining purchasing behavior is the tendency, varying in strength, to adapt to the purchase situation. This particular point is underlined by several authors. Researchers often wrongly consider that the purchase is made as soon as the consumer has decided to go ahead with it. Bagozzi and Warshaw (1990) propose, on the contrary, that purchasing behaviors should be seen as tests or attempts, which sometimes come up against obstacles, linked to the purchase situation. For example, stockouts in a store or the unavailability of credit can thwart consumers’ intentions. Oliver (1999) also emphasizes the need to take these obstacles into account when one thinks about defining loyalty. For Oliver (1999) the loyal consumer is someone who is aware of these obstacles and tries to overcome them. We take up this idea in our theoretical framework by including in each case consumers’ tendency to control the situations that arise. In sum, the eclectic framework that we put forward brings together theories that other authors have very often tended to reject. It also incorporates proposals for improving the concept of loyalty found in the literature (Figure 1C). As Jacoby and Kyner (1971) suggest, we define loyalty both in terms of its consequences – observed behaviors – and its causes, which encourage customers to control the evolution of their initial purchase level. But our framework enriches the definition offered by Jacoby and Kyner (1971). It specifies their condition number 3 (i.e. “expressed over time”), by opting for the long term. This orientation is characteristic of reward programs, which require their users to accumulate points over a long period and encourage them to maximize their utility in the long rather than short term, on future rather than immediate purchases (Kopalle and Neslin 2003). Finally, consumers’ desire to surmount the obstacles that make purchasing more difficult (Oliver 1999) is taken into account by their tendency to control these obstacles in any given purchasing situation. We therefore define the tendency to loyalty behavior as the customer’s tendency to maintain or increase, from one period to the next, his initial expenditure level, despite the obstacles encountered.

Research hypotheses Current definitions of loyalty programs clearly situate the origin of their effects in the rewards and incentives they procure for their users (Yi and Jeon 2003; Liu 2007). The great majority of researchers explain these effect by means of the theory of effective learning (Nord and Peter 1980; Rothschild and Gaidis 1981; Foxall 1996, 1997;

Journal of Marketing Trends - Volume I (February 2011) 37


24366_int_Mise en page 1 22/12/10 12:25 Page38

J O U R N A L O F M A R K E T I N G T R E N D S - L O YA LT Y P R O G R A M

notre cadre enrichit la définition de Jacoby et Kyner (1971). Il précise leur condition numéro 3 (i,e, « exprimée au cours du temps » ), en optant pour le long terme. Cette orientation est celle des programmes de récompenses. Ceux-ci imposent en effet aux utilisateurs d’accumuler des points sur une longue période et les poussent à maximiser leur utilité sur le long plutôt que sur le court terme, sur les achats à venir plutôt que sur les achats immédiats (Kopalle et Neslin 2003). Enfin, le désir du consommateur de surmonter les obstacles qui rendent plus difficile l’achat (Oliver 1999) est pris en compte par sa tendance à contrôler ces obstacles dans chaque situation d’achat donnée. Nous définissons donc la tendance au comportement de fidélité comme la tendance du client à maintenir ou à augmenter, de période en période, son niveau initial de dépenses, malgré les obstacles qu’il rencontre.

Hypothèses de recherche Les définitions courantes des programmes de fidélité situent clairement l’origine de leurs effets dans les récompenses et les incitations qu’ils procurent à leurs utilisateurs (Yi et Jeon 2003 ; Liu 2007). Dans leur grande majorité les chercheurs expliquent ces effets par la théorie de l’apprentissage opérant (Nord and Peter 1980 ; Rothschild et Gaidis 1981 ; Foxall 1996, 1997 ; Sharp and Sharp 1997 ; Keh et Lee 2006 ; Liu 2007). Les points constituent des récompenses indirectes, ayant une fonction de renforçateurs secondaires (RS). Ils renforcent le comportement d’achat parce que le client qui les reçoit en achetant les associe aux récompenses tangibles futures. Les réductions, ou rabais obtenus en échange des points gagnés jouent le rôle de renforçateurs primaires (RP). Ils renforcent le comportement d’achat en faveur du distributeur qui remet les récompenses, car le client associe ces récompenses au distributeur qui les lui donne. Cependant deux cas sont à considérer . Dans le premier on considère les clients exclusivement fidèles à deux enseignes X et Y en concurrence, mais qui sont partenaires d’un programme multi-sponsors (Figure 2 A). Les points (RSX ou RSY) sont octroyés par le partenaire X ou Y et échangé par le client, respectivement auprès de X ou de Y, contre des récompenses tangibles. Les renforcements du comportement d’achat du client (RPX ou RPY) et (RSX ou RSY), bénéficient intégralement à X ou à Y. Il n’y a pas par exemple de transfert possible des renforcements secondaires RSX vers les renforcement primaires, c’est à dire RSX = RPX et RSY = RPY si tous les points sont utilisés dans les délais. Ce programme fonctionne alors comme un programme mono-sponsor (figure 2 A).

38 ISSN 1961-7798 - © 2009, International Marketing Trends Conference

Sharp and Sharp 1997; Keh and Lee 2006; Liu 2007). Points are indirect rewards, having a secondary reinforcement (SR) function. They reinforce purchasing behavior because customers who receive them through their purchases associate them with future tangible rewards. Reductions, or discounts obtained in exchange for points gained, play the role of primary reinforcers (PR). These reinforce purchasing behavior in favor of the retailer which provides the rewards, since customers associate these rewards with that particular retailer. There are, however, two cases to consider. In the first, one is concerned with customers who are loyal exclusively to two competing retailers X and Y, but which are partners in a multi-sponsor program (Figure 2A). The points (PRX or PRY) are bestowed by partner X or Y and exchanged by the customer, at X or Y respectively, for tangible rewards. The reinforcers of the customer’s purchasing behavior (PRX or PRY) fully benefit either X or Y. There is, for example, no possible transfer of secondary reinforcers (SRX) to primary reinforcers, i.e. SRX = PRX and SRY = PRY if all the points are used within the validity period. This program thus functions like a single-sponsor program (Figure 2A). Figure 2. Impact of gratification on purchasing behavior.

In the second case we consider multi-loyal customers in the same program (Figure 2B). The purchases made by these customers at X or Y generate points, whose reinforcer effects on purchasing (SRX or SRY) proportionally benefit the sales of the retailers X and Y respectively. There is, however, a difference in relation to the customers who are exclusive to these programs. Multi-loyal customers can choose to obtain the refund on their points with either of the competing stores X and Y. Points gained at X can be used to obtain rewards at Y and vice versa. The benefit of the primary reinforcers (PR) no longer


24366_int_Mise en page 1 22/12/10 12:25 Page39

J O U R N A L O F M A R K E T I N G T R E N D S - L O YA LT Y P R O G R A M

Figure 2. Impact of gratification on purchasing behavior.

Dans le second cas on considère les clients multifidèles du même programme (Figure 2 B). Les achats réalisés par ces clients chez X et chez Y, génèrent des points dont les effets de renforcement sur l’achat (RSX ou RSY), bénéficient respectivement au pro rata des achats, aux distributeurs, X et Y. Toutefois il existe une différence par rapport aux clients exclusifs de ces programmes. Les clients multi-fidèles peuvent choisir d’obtenir le remboursement de leurs points auprès des enseignes concurrentes X ou Y de leur choix. Des points gagnés chez X peuvent procurer des récompenses chez Y et vice versa. Le bénéfice des renforcements primaires (RP), ne dépend plus seulement du programme mais aussi de l’arbitrage du client. Et comme par définition le client n’est pas fidèle de façon exclusive, chaque enseigne peut redouter que son choix lui soit défavorable. Mais de manière plus objective on peut aussi penser que la multi-fidélité dénote un comportement opportuniste. Il amènera le client à prendre ses récompenses tangibles de façon plus équilibrée, tantôt chez X tantôt chez Y, selon les circonstances du moment. L’évolution des comportements d’achat des clients utilisant un programme multisponsors, ne devraient donc pas être très différents d’une enseigne en concurrence à l’autre. Cette remarque vaut aussi bien pour les clients exclusifs, n’achetant que dans une seule enseigne, que pour les clients multi-fidèles achetant régulièrement dans les deux enseignes. D’autre part, l’évolution du comportement d’achat des clients multi-fidèles dans l’une des deux enseignes en concurrence, ne devrait pas se faire au détriment de l’autre. Les achats dans les deux enseignes procurent en effet les mêmes renforcements et le programme n’induit pas de biais a priori dans les choix des clients. D’où les trois hypothèses de recherche suivantes.

depends solely on the program but also on the customer’s decision. And since by definition customers are n o t exclusively loyal, each company may fear their choices may be unfavorable to it. But in a more objective way, one may also think that multi-loyalty indicates opportunistic behavior. It will lead the customer to take tangible rewards in a more balanced way, sometimes at X and sometimes at Y, according to the circumstances of the moment. The evolution of the purchasing behaviors of customers using a multi-sponsor program should not therefore be very different from one competing company to the other. This comment also applies as much to exclusive customers, who shop at only one of the stores, as to multi-loyal customers who regularly shop at both. On the other hand, the evolution of the purchasing behavior of multi-loyal customers in one of the two competing stores should not adversely effect the other. Indeed purchases in the two stores produce the same reinforcers and the program does not lead to any inherent bias in customers’ choices. Hence the following three research hypotheses. H1 The tendencies of exclusively loyal users in regard to the initial purchase and to control of purchasing have the same distribution in the two competing stores. H2 The tendencies of multi-loyal users in regard to the initial purchase and to control of purchasing have the same distribution in the two competing stores. H3 The tendency of multi-loyal users to control the evolution of their purchases in favor of one store is independent of their tendency to control the evolution of their purchases with the other competing store.

METHOD Deployment of concepts The deployment of the concepts of latent initial purchase tendency, latent tendency to control the initial purchase, and latent tendency to adapt to purchasing situations is implemented by means of a latent growth measurement model. This is briefly described here but a more detailed account is provided by Frisou (2005). The measurement model of repeat purchase behavior comprises two latent factors (Figures 3A and 3B).

Journal of Marketing Trends - Volume I (February 2011) 39


24366_int_Mise en page 1 22/12/10 12:25 Page40

J O U R N A L O F M A R K E T I N G T R E N D S - L O YA LT Y P R O G R A M

H1 Les tendances à l’achat initial et au contrôle de l’achat, des utilisateurs exclusivement fidèles, ont des distributions identiques dans les deux enseignes en concurrence.

Figure 3. Model of latent trajectories

H2 Les tendances à l’achat et au contrôle de l’achat, des utilisateurs multi-fidèles, ont des distributions identiques dans les deux enseignes en concurrence. H3 La tendance des utilisateurs multi-fidèles à contrôler l’évolution de leurs achats en faveur d’une enseigne, est indépendante de leur tendance à contrôler l’évolution de leurs achats avec l’autre enseigne en concurrence.

MÉTHODOLOGIE Mise en œuvre des concepts La mise en œuvre des concepts de tendance latente à l’achat initial, au contrôle de l’achat initial, et à l’adaptation aux situations d’achat, est réalisé au moyen d’un modèle de mesure de croissance latente. Ce modèle est succinctement décrit ici mais une présentation plus détaillée est présentée par Frisou (2005). Le modèle de mesure du comportement d’achat répété comprend deux facteurs latents (figures 3A et 3B).

These are interpreted as chronometric common factors. The latent factor α represents the initial purchase level of customers during the first period “t = 0”. The latent factor β expresses the latent tendency of customers to change the initial level of expenditure from one period to the next. The observed purchase amount “Ait” for customer “i”, in each period “t”, is expressed by equation (1): (1) Αit = αi + βi ∗ t + εit

Figure 3. Model of latent trajectories

In the first period t = 0 and equation (1) takes the following form (2): (2) Ai0 = αi + εi0 ,

Equations (3) and (4) which follow, complete the specification of the measurement model and introduce the variances of the chronometric factors. (3) αi = E(α) + ζ αi Ils sont interprétés comme des facteurs communs chronométriques. Le facteur latent α représente le niveau initial d’achat des clients durant la première période « t = 0 ». Le facteur latent β exprime la tendance latente des clients à faire évoluer de période en période le niveau initial de dépenses α. D'un point de vue mathématique α est une ordonnée à l’origine et β une pente. Le montant d’achat observé « Ait » pour le client « i », à chaque période « t », est exprimé par l’équation 1: (1) Αit = αi + βi ∗ t + εit Dans la première période t = 0 l’équation (1) prend la forme (2) suivante : (2) Ai0 = αi + εi0 , Les équations (3) et (4) qui suivent, complètent la spécification du modèle de mesure et introduisent les moyennes et les variances des facteurs chronométriques. (3) αi = E(α) + ζ αi (4) βi = E(β) + ζ βi

40 ISSN 1961-7798 - © 2009, International Marketing Trends Conference

(4) βi = E(β) + ζ βi

Testing the hypotheses Data collection The data used are taken from a multi-store loyalty program, created by partner retailers in a French town of 22,000 inhabitants. The multi-store program, which began at the end of 2005, comprises 29 retailers. It currently has 9150 customers with the loyalty cards, of whom 64% live in the town center, 15% live with 20 km of the center, 9% in the département and 12% outside the département. We were interested in the customers of two sports goods franchise stores, representing two of the leading sport goods companies in Europe. To preserve the anonymity of these companies we shall call them respectively “Planet Sport” (PS) and “Sport Passion” (SP). The cohort of card-holders which was followed is composed of customers of these two stores who joined the program in the 7 months following its launch, i.e. from December


24366_int_Mise en page 1 22/12/10 12:25 Page41

J O U R N A L O F M A R K E T I N G T R E N D S - L O YA LT Y P R O G R A M

Mise à l’épreuve des hypothèses Collecte des données Les données utilisées sont issues d’un programme de fidélité multi-enseignes, créé par l’association des commerçants d’une ville française de 22 000 habitants. Ce programme multi-enseignes qui a débuté fin 2005 réunit 29 commerçants, Aujourd’hui il compte 9150 clients porteurs de carte dont 64 % résident en centre ville, 15 % habitent à moins de 20 km du centre, 9 % dans le département et 12% hors département. Nous nous sommes intéressés aux clients de deux magasins franchisés d’articles de sport, représentant deux grandes enseignes qui occupent les deux premières places en Europe. Pour préserver l’anonymat de ces sociétés nous les appellerons respectivement « Planète Sport » (PS) et « Sport Passion » (SP). La cohorte de porteurs qui a été suivie est constituée des clients de ces deux enseignes qui ont rejoint le programme dans les 7 premiers mois de son lancement, c’est à dire de décembre 2005 à juin 2006. Elle est composée de 1952 porteurs de cartes, dont 965 sont exclusivement fidèles à Planète Sport, 239 sont exclusivement fidèles à Sport Passion et 748 achètent dans les deux enseignes avec leur carte. Durant ces deux premières années 62 % des clients sont restés fidèles à l’une des deux enseignes alors que 38 % ont manifesté un comportement de multi-fidélité. Le nombre total de transactions réalisées par les clients dans les deux enseignes, du 2° semestre 2006 au 1° semestre 2008, s’élève à 14 288 achats, tous récompensés par des points de fidélité. Les règles d’attribution des points sont identiques dans les deux magasins qui sont tous les deux implantés au cœur de la ville. Les clients de cette cohorte ont réalisé une moyenne de 8,37 achats dans ces deux enseignes en deux ans avec un maximum individuel de 22 achats.

Test de l’hypothèse H1 L’hypothèse H1 soutient que les utilisateurs exclusivement fidèles à chacune des deux enseignes participant au programme, ont des tendances identiques à acheter et à contrôler leur niveau d’achat initial. En d’autres termes les distributions de ces tendances sont les mêmes chez les clients exclusifs des deux enseignes en concurrence. Un modèle de croissance latente comportant deux groupes a donc d’abord été estimé. Le premier groupe concerne les porteurs de cartes exclusivement fidèles à Planète Sport (965 individus), le second groupe a trait aux porteurs de cartes exclusivement fidèles à Sport Passion (239 individus). Les résultats de ce modèle, reportés dans le tableau 1, font apparaître des indices d’ajustement qui traduisent une bonne adéquation des données au modèle. Les tendances latentes moyennes à l’achat initial sont sensiblement différentes pour les deux groupes.33,18 ⇔ pour les clients de Sport Passion contre seulement 20,76 ⇔ pour les clients de Planète Sport. Mais les tendances latentes moyennes au contrôle de l’achat initial sont très voisines, - 2,30 ⇔ pour les clients de Planète Sport, contre - 2,01 ⇔

2005 to June 2006. The cohort is composed of 1952 card holders, of whom 965 are exclusively loyal to Planet Sport, 239 are exclusively loyal to Sport Passion and 748 make purchases in both stores with their card. During the first two years, 62% of customers remained loyal to one or other of the two stores, while 38% manifested multi-loyal behavior. The total number of purchases made in the two stores, from the 2nd half-year 2006 to the 1st half-year 2008, was 14,288, all of which were rewarded with loyalty points. The rules for allocating points is the same in the two stores, both of which are situated in the town center. The customers in this cohort made an average of 8.37 purchases in these two stores over two years, with an individual maximum of 22 purchases.

Testing hypothesis H1 Hypothesis H1 proposes that users who are exclusively loyal to each of the two stores participating in the program have the same tendency to purchase and control their initial purchasing level. In other words, the distribution of these tendencies are the same for customers exclusive to the two competing companies. A latent growth model comprising two groups has therefore to be worked out first. The first group contains card holders who are exclusively loyal to Planet Sport (965 individuals), and the second group contains card holders who are exclusively loyal to Sport Passion (239 individuals). The results of this model, shown in Table 1, reveal goodness-of-fit indices that show a good match of the data to the model. The mean latent initial purchase tendencies are markedly different for the two groups: ⇔33.18 for Sport Passion customers against only ⇔20.76 for Planet Sport customers. But the mean latent tendency to control the initial purchase are very similar: – ⇔2.30 for Sport Passion customers against –⇔2.01 for Planet Sport customers. If this last mean is not significant (t = –0.77), this is due to small size of the sample (239). Table 1 – Model 1 – Latent trajectories of customers exclusively loyal to each store

The mean latent purchasing trajectories of customers exclusive to the two stores can be seen from the graph in Figure 4.

Journal of Marketing Trends - Volume I (February 2011) 41


24366_int_Mise en page 1 22/12/10 12:25 Page42

J O U R N A L O F M A R K E T I N G T R E N D S - L O YA LT Y P R O G R A M

pour les clients de Sport Passion. Si cette dernière moyenne n’est pas significative ( t = -0,77), cela est dû à la taille réduite de l’échantillon (239).

Figure 4 – Mean latent trajectories of customers loyal to the two stores

Table 1 – Model 1 – Latent trajectories of customers exclusively loyal to each store

Les trajectoires d’achat latentes moyennes, des clients exclusifs des deux enseignes peuvent être visualisées sur le graphique reproduit en figure 4 Figure 4 – Mean latent trajectories of customers loyal to the two stores

Mettre à l’épreuve l’hypothèse H1, revient à tester l’invariance des moyennes et des variances des facteurs chronométriques latents à travers les deux groupes. Ce test suppose de construire un ensemble hiérarchisé de modèles « nichés », qui évoluent pas à pas vers les prédictions de l’hypothèse H1. Chaque modèle est ensuite comparé à son prédécesseur à partir de la différence de Chi-deux entre les deux modèles (Bollen 1989). Le seuil de signification choisi est p < 0,01. Nous aurions pu choisir p < 0,05 ou p < 0,10. Nous justifions ce seuil assez bas par le fait que la valeur du chi-deux dépend de la taille des échantillons. De grands échantillons tels que le notre favorisent donc le rejet des hypothèses testées. Le premier modèle H forme, estime librement les paramètres λβ3 et λβ4 dans les deux groupes. Les trajectoires sont donc supposées de formes quelconques. Le χ² et sa probabilité associée (p = 0,72) témoignent d’un excellent ajustement. Table 2 – Analysis of invariance of exclusive customers’ latent trajectories

42 ISSN 1961-7798 - © 2009, International Marketing Trends Conference

Testing hypothesis H1 amounts to testing the invariance of means and the variance of latent chronometric factors across the two groups. This test presupposes constructing a hierarchized set of “nested” models, which gradually evolve toward the predictions of hypothesis H1. Each model is then compared with its predecessor from the chi-square difference between the two models (Bollen 1989). The significance threshold chosen is p < 0.01. We could have chosen p < 0.05 or p < 0.10. We justify this rather low threshold by the fact the value of chi-square depends on the sample size. Large samples such as ours therefore favor the rejection of the tested hypotheses. The first model Hforme freely estimates the parameters λβ3 and λβ4 in the two groups. The trajectories are therefore assumed not to be standard forms. The χ² and its associated probability (p = 0.72) indicate an excellent goodness-of-fit. Table 2 – Analysis of invariance of exclusive customers’ latent trajectories

The second model (HΛ) constrains these parameters to equality across the two groups. The chi-square difference test that compares the second model to the first does not result in this hypothesis being rejected (p = 0.782). In the third model we put forward the further hypothesis that the trajectories are linear. Here too the chi-square difference test does not lead to the rejection of the model (p = 0.036). The fourth model (termed HΛλ μ) adds to the previous one the hypothesis of equality of tendency means across the groups. Again the chi-square difference test does not lead to its rejection (p = 0.016) This is also the case for the fifth model (termed HΛλμ ψ), which adds the hypothesis of


24366_int_Mise en page 1 22/12/10 12:25 Page43

J O U R N A L O F M A R K E T I N G T R E N D S - L O YA LT Y P R O G R A M

Le deuxième modèle (HΛ) contraint ces paramètres à l’égalité à travers les deux groupes. Le test de différence de chi-deux qui compare le second modèle au premier nous conduit à ne pas rejeter cette hypothèse (p = 0,782). Dans le troisième modèle on fait l’hypothèse supplémentaire que les trajectoires sont linéaires. Le test de différence de chi-deux ne conduit pas au rejet du modèle (p = 0,036). Le quatrième modèle (noté HΛλ μ) ajoute au précédent l’hypothèse d’égalité des moyennes des tendances à travers les groupes. Le test de différence de chi-deux ne conduit pas non plus à son rejet (p = 0,016) C’est également le cas pour le cinquième modèle (noté HΛλμ ψ ), qui ajoute l’hypothèse d’égalité des variances des tendances à travers les groupes (p = 0,016). Seul le sixième modèle, adjoignant l’hypothèse d’égalité des variances résiduelles à travers les deux groupes est rejeté (p = 0,000). Au terme de cette analyse hiérarchique on convient donc d’accepter le cinquième modèle qui corrobore l’hypothèse H1. Les trajectoires latentes d’achat des porteurs de cartes, dont la fidélité envers chacune des enseignes est exclusive sont comparables.

Test de l’hypothèse H2 Dans l’hypothèse H2 il est affirmé que les tendances à l’achat et au contrôle de l’achat, des utilisateurs multifidèles, ont dans les deux enseignes en concurrence des distributions comparables. La mise à l’épreuve de cette hypothèse est effectuée de manière analogue à la précédente. Un modèle de croissance latente comportant deux groupes est estimé. Le premier groupe a trait au comportement d’achat des 748 clients multi-fidèles chez Planète Sport. Le second groupe concerne les comportement d’achats des mêmes 748 clients multi-fidèles chez Sport Passion. Les résultats de l’estimation de ce modèle sont reproduits dans le tableau 3. Les indices d’ajustement traduisent dans leur ensemble une assez bonne adéquation des données au modèle. On note un quasi parallélisme de l’évolution du comportement tendanciel d’achat moyen (figure 5). Figure 5 – Mean latent trajectories of customers of the two stores

equality of tendency variances across the groups (p = 0.016). Only the sixth model, adding the hypothesis of equality of residual variances across the two groups, is rejected (p = 0.000). Following this hierarchical analysis it is therefore appropriate to accept the fifth model, which confirms hypothesis H1. The latent purchasing trajectories of card-holders, with loyalty toward each of the stores is exclusive, are comparable.

Testing hypothesis H2 In hypothesis H2 it is asserted that the purchasing and purchase control tendencies of multi-loyal users have comparable distributions in the two competing retail outlets. This hypothesis was tested in a similar way to the previous hypothesis. A latent growth model comprising the two groups was estimated. The first group is related to the purchasing behavior of 748 multi-loyal customers at Planet Sport. The second group concerns the purchasing behavior of the same 748 multi-loyal customers at Sport Passion. The results of assessing this model are shown in Table 3. The goodness-of-fit indices as a whole reflect a relatively good match of the data to the model. We see that the development of the mean purchasing tendency behavior is very closely matched (Figure 5). Figure 5 – Mean latent trajectories of customers of the two stores

The mean latent tendencies of the initial purchase are very close in the two groups: 52.57 and 50.10. They are also some 20 euros higher than those of customers who are exclusively loyal to one or other of the two companies. Multi-loyal customers are probably more experienced and therefore more vulnerable to competition. This finding seems to be confirmed in the mean latent tendencies of the initial purchase control which are negative and similar to each other: –7.59 as against –7.42. Logically enough, we see that these gradients are steeper than those of exclusively loyal customers.

Les tendances latentes moyennes à l’achat initial sont très proches dans les deux groupes 52,57 versus 50,10. Elles sont aussi supérieures d’une vingtaine d’euros à

Journal of Marketing Trends - Volume I (February 2011) 43


24366_int_Mise en page 1 22/12/10 12:25 Page44

J O U R N A L O F M A R K E T I N G T R E N D S - L O YA LT Y P R O G R A M

celles des clients fidèles exclusivement aux deux enseignes. Les clients multi-fidèles sont sans doute plus expérimentés et donc plus vulnérables à la concurrence. Cette remarque semble se confirmer dans les tendances latentes moyennes au contrôle de l’achat initial qui sont négatives et voisines l’une de l’autre, –7,59 versus –7, 42. On constate assez logiquement que ces pentes sont plus accentuées que celles des clients dont la fidélité est exclusive.

Table 3 – Model 2 – Latent trajectories of customers multi-loyal to the two stores

Table 3 – Model 2 – Latent trajectories of customers multi-loyal to the two stores

La mise à l’épreuve de l’hypothèse H2 est réalisée comme celle de l’hypothèse H1 (tableau 4). L’étude de l’invariance des paramètres dans les deux groupes montre des résultats similaires. La comparaison des modèles nichés, qui correspondent à une hiérarchie d’hypothèses de plus en plus contraignante, nous conduit à accepter tous les modèles y compris le modèle 6 (p = 0,600) On ne rejette donc pas le sixième modèle qui valide l’hypothèse H2. Les trajectoires latentes d’achat des porteurs de cartes fidèles aux deux enseignes sont comparables.

Hypothesis H2 was tested in the same way as hypothesis H1 (Table 4). Analysis of the invariance of parameters in the two groups shows similar results. The comparison of the nested models, which correspond to an increasingly restrictive hierarchy of hypotheses, results in the acceptance of all the models including model 6 (p = 0.600) Therefore we do not reject model 6, which confirms hypothesis H2. The latent purchasing tendencies of card-holders who are loyal to one or other of the two stores are comparable. Table 4 – Analysis of invariance of multi-loyal customers’ latent trajectories

Table 4 – Analysis of invariance of multi-loyal customers’ latent trajectories

Testing hypothesis H3

Test de l’hypothèse H3 L’hypothèse H3 soutient que les clients multi-fidèles, ont des tendances au contrôle de l’achat initial dans chacune des enseignes, qui sont indépendantes. Autrement dit, l’évolution des achats réalisés avec la carte de fidélité auprès d’une enseigne ne se fait pas au détriment de l’autre enseigne. Le test de cette hypothèse a été effectué en considérant la corrélation entre ces deux tendances. Le modèle support est le modèle de croissance latente comportant les deux processus d’achat, dont le path diagram est reproduit en figure 6.

44 ISSN 1961-7798 - © 2009, International Marketing Trends Conference

Hypothesis H3 maintains that multi-loyal customers have initial purchase control tendencies in the two stores that are independent of each other. In other words, the evolution of purchases made with the loyalty card at one store is not to the detriment of the other store. This hypothesis was tested by looking at the correlation between these two tendencies. The model is the latent growth model composed of the two purchasing processes, the path diagram of which is shown in Figure 6.


24366_int_Mise en page 1 22/12/10 12:25 Page45

J O U R N A L O F M A R K E T I N G T R E N D S - L O YA LT Y P R O G R A M

Figure 6 – Path diagram of behavioral model of multi-loyal customers

Figure 6 – Path diagram of behavioral model of multi-loyal customers

L’estimation de ce modèle avec l’estimateur robuste du maximum de vraisemblance, présente une assez bonne qualité d’ajustement. La valeur du χ² est de 10,966 avec 5 degrés de liberté ( p = 0, 052). CFI = 0,89, TLI = 0,85 RMSEA = 0,04 et SRMR = 0,06 sont dans leur ensemble à des niveaux jugés acceptables dans la littérature.

Estimation of this model with the robust estimator of maximum likelihood shows a relatively good fit. The value of χ² is 10.966 with 5 degrees of freedom (p = 0.052). CFI = 0.89, TLI = 0.85 RMSEA = 0.04 and SRMR = 0.06 are generally considered acceptable in the literature.

La corrélation entre les tendances latentes à faire évoluer le niveau initial d’achat chez Sport Passion et chez Planète Sport est estimée à 0,11. Cette très faible valeur n’est pas significativement différente de zéro ( t = 1,18). Un modèle niché dans lequel la valeur de cette corrélation est fixée à zéro, n’est pas rejeté par le test de différence de Chi-deux approprié (Δχ² = 1,499 avec 1 degré de liberté et p = 0,22).L’hypothèse H3 est donc aussi vérifiée. En revanche la corrélation entre les tendances à l’achat initial chez Sport Passion et chez Planète Sport est significative (t = 2,72) et estimée à – 0, 19. En d’autres termes, plus un client a tendance en début de programme à acheter chez Sport Passion, plus il a tendance à acheter chez Planète Sport et réciproquement. Les clients multi-fidèles présentent non seulement une tendance à l’achat initial plus élevée que les clients exclusifs, mais ils semblent aussi acheter de manière complémentaire et non concurrente dans les deux enseignes.

CONCLUSION Au terme de cette recherche il convient de conclure en soulignant ses limites, ses apports et les contributions qu’elle apporte sur le plan managérial.

Limites et recherches futures Ses limites sont essentiellement méthodologiques. Nous avons par exemple évoqué le concept de tendance

The correlation between the latent tendencies to change the initial purchase level at Sport Passion and at Planet Sport is estimated to be 0.11. This very low value is not significantly different from zero (t = 1.18). A nested model in which the value of this correlation is set at zero is not rejected by the appropriate chi-square difference test (Δχ² = 1.499 with 1 degree of freedom and p = 0. 22). Hypothesis H3 is therefore also confirmed. On the other hand, the correlation between the initial purchase tendencies at Sport Passion and Planet Sport is significant (t = 2.72) and estimated to be – 0.19. Phrased differently, the more a customer tends to shop at Sport Passion at the beginning of the program, the more he will tend to shop at Planet Sport, and vice versa. Multi-loyal customers display not only a higher initial purchase tendency than exclusive customers, but they also seem to buy in a complementary and non-competitive way in the two stores.

CONCLUSION In conclusion to this study we should emphasize its limitations, contributions and managerial implications.

Limitations and Further Research Its limitations are largely methodological. For example, we have drawn on the concept of tendency to loyalty and we have measured this by the latent gradient given by the purchasing process. This loyalty is loyalty to the program, not to the retailers such as Sport Passion

Journal of Marketing Trends - Volume I (February 2011) 45


24366_int_Mise en page 1 22/12/10 12:25 Page46

J O U R N A L O F M A R K E T I N G T R E N D S - L O YA LT Y P R O G R A M

à la fidélité et nous l’avons mesuré par la pente latente qui commande le processus d’achat. Cette fidélité est une fidélité au programme, et non aux commerçants qui en sont membres comme Sport Passion et Planète Sport. En effet certains clients porteurs de la carte ne l’utilisent pas ou peu. Ils continuent d’acheter dans ces commerces sans bénéficier des points et des récompenses, témoignant ainsi d’une fidélité désintéressée aux enseignes. En revanche leurs achats ne sont pas enregistrés et ces clients sont probablement considérés comme non fidèles. Fidélité au commerce et fidélité au programme ne coïncident donc pas. Une autre limite plus générale a trait au recueil des données et la représentativité des échantillons. Ces échantillons reprennent l’intégralité des individus des cohortes étudiées, mais n’en restent pas moins des échantillons de convenance. La validité externe de cette recherche suppose de la dupliquer sur d’autres programmes et d’autres types de commerce. Les résultats que nous enregistrons avec des enseignes commercialisant des articles de sport très particuliers, peuvent-ils être observés dans le cas de librairies ou de boulangeries dont les produits sont moins différenciés ? Ce qui est valable dans les espaces du petit commerce de détail l’est–il aussi dans les hypermarchés ? La duplication de cette recherche dans des contextes d’achat et de types de programme très différent s’avère nécessaire.

Apports et implications managériales Ces limites étant précisées, il reste que ce travail dégage un apport théorique important qui répond a des préoccupations managériales essentielles. Le programme de fidélité multi-sponsors étudié suscite auprès des clients des comportements d’achat similaires d’une enseigne concurrente à l’autre. Si la validité externe de cette recherche était démontrée, le champ d’application des programmes « mutualisés » s’agrandirait considérablement. Réservés jusqu’alors aux enseignes distribuant des biens et des services différents et complémentaires, ces programmes pourraient associer des partenaires en concurrence directe. Cette perspective est du plus grand intérêt pour les gestionnaires des programmes multienseignes associant les grandes marques, tels que S’Miles, Maximiles, ou Mouvango. Mais elle l’est aussi pour les managers de programmes de fidélité travaillant à la revitalisation des centres urbains. Il est en effet possible d’envisager des associations de commerces dans un centre ville, sans être freiné par le fait que leur offre est identique. Il est aussi possible de concevoir des programmes de revitalisation des centres urbains communs à plusieurs villes voisines qui se partagent une clientèle de plus en plus mobile. La compétition ne serait donc pas l’ennemi de la fidélisation.

46 ISSN 1961-7798 - © 2009, International Marketing Trends Conference

and Planet Sport who are members of it. Indeed, some card-holding customers do not use it all or do so sparingly. They continue purchasing in these stores without benefiting from points or rewards, thus testifying to a disinterested loyalty to the stores. On the other hand their purchases are not recorded and these customers are probably considered to be non-loyal. Loyalty to the stores and loyalty to the program do therefore not coincide. Another more general limitation concerns the data collection and the representativeness of the samples. Our samples contain all the individuals in the cohorts studied, but they nonetheless remain convenience samples. The external validity of this study counts on it being reproduced on other programs and other types of retailing. Can the results we obtained from stores selling highly specialized sports equipment also be found for, say, bookstores or bakers, where the products are less differentiated? Is what is valid for small-scale retail outlets also valid for hypermarkets? The reproduction of this study in very different shopping contexts and types of loyalty programs is thus undoubtedly called for.

Findings and Managerial Implications Despite these limitations, this study nonetheless makes an important theoretical contribution that answers to some basic managerial concerns. The multi-sponsor loyalty program studied gives rise to very similar purchasing behaviors on the part of customers in the two competing stores. If the external validity of this study were to be demonstrated, the field of application of such “mutualized” programs would be considerably enlarged. Limited until now to stores retailing goods and services that are different and complementary, these programs could bring together partners in direct competition with each other. This prospect is of greater interest for multi-company programs that bring together large brands such as S’Miles, Maximiles or Mouvango. But it is also of potential interest to loyalty program managers working toward the revitalization of city centers. Indeed it is possible to envisage associations of retailers in the city centre, without being held back by the fact that their offering may the same. It is also possible to conceive of programs for revitalizing urban centers involving several neighboring towns which share a more or less mobile clientele. Competition would therefore not be opposed to loyalty creation. Co-marketing has accustomed us to brand alliances in different stages of product development (shared development, co-branding, joint advertising). Co-loyalty creation, which emerges from multi-sponsor loyalty programs, is also inspired by the same idea of widening the retailer’s territory. Like co-marketing, it also has its


24366_int_Mise en page 1 22/12/10 12:25 Page47

J O U R N A L O F M A R K E T I N G T R E N D S - L O YA LT Y P R O G R A M

Le co-marketing nous a habitué aux alliances de marques dans les différents stades de développement du produit (développement partagé, co-marquage, communication conjointe). La co-fidélisation, qui se met en place derrière les programmes de fidélité multi-sponsors, s’inspire elle aussi de la même idée d’élargissement du territoire du distributeur. Comme le co-marketing elle comporte elle aussi des risques. En particulier les risques de dilution de l’enseigne ou de cannibalisation des ventes. Risques, que des recherches telles que la notre tentent d’évaluer le plus précisément possible.

risks, in particular those of retail brand dilution or the cannibalization of sales. They are risks that studies such as ours try to assess as accurately as possible.

Bibliographie / Bibliography Allaway, Arthur W., Richard M. Gooner, David Berkowitz and Lenita Davis (2006), “Deriving and Exploring Behavior Segments Within a Retail Loyalty Card Program”, European Journal of Marketing, 40, Issue 11/12, 1317-1339. Amine, Abdelmajid (1998), “Consumers’ True Brand Loyalty: the Central Role of Commitment”, Journal of Strategic Marketing , 6, 305-319. Bagozzi, Richard P. and Paul R. Warshaw (1990), “Trying To Consume”, Journal of Consumer Research, 17 (September), 127-140. Bollen, Kenneth A. (1989), Structural Equations With Latent Variables, New York, John Wiley & Sons. Coutu, Walter (1949), Emergent Human Nature: A Symbolic Field Interpretation, Alfred A. Knopf ed. New York. Dall’ Olmo Riley, Francesca., Andrew.S.C Ehrenberg., Stephen B. Castleberry, Patrick Barwise and Neil. R. Barnard (1997), “The variability of attitudinal repeat-rates”, International Journal of Research in Marketing, 14, 5, 437-450. Day, George S. (1969), “A Two Dimensional Concept of Brand Loyalty”, Journal of Advertising Research, 9, 3, 29-35. Dick, Alan S. and Kunal Basu (1994), «Customer Loyalty: Toward an Integrated Conceptual Framework,» Journal of the Academy of Marketing Science, 22 (2), 99-113. Dowling Grahame R. and Mark Uncles (1997), “Do Customer Loyalty Programs Really Work ?”, Sloan Management Review, (Summer), 71-82. Ehrenberg, Andrew. S. C. (2000). «Repeat Buying - Facts, Theory and Applications.» Journal of Empirical Generalisations in Marketing Science, 5, 392-770. Filser, Marc (1994), Le comportement du consommateur, Paris, Dalloz. Foxall, Gordon (1996), Consumer in Context: The BPM Research Program, Routdlege , London. Foxall, Gordon (1997), Psychology Marketing, Macmillan Press LTD, London.

Journal of Marketing Trends - Volume I (February 2011) 47


24366_int_Mise en page 1 22/12/10 12:25 Page48

J O U R N A L O F M A R K E T I N G T R E N D S - L O YA LT Y P R O G R A M

Frisou, Jean (2005), “A Trend Approach of Loyalty Behavior: from the Concept to its Measure”, Recherche et Applications en Marketing, 20, 2, 105-125. Jacoby, Jacob and David Kyner (1973), “Brand Loyalty vs Repeat Purchasing Behavior”, Journal of Marketing Research, 10 (February), 1-9. Kapferer, Jean-Noël and Laurent Gilles. (1983), La sensibilité aux marques: un nouveau concept pour gérer les marques, Fondation Jours de France pour la Recherche en Publicité. Keh, Hean Tat and Yih Hwai Lee (2006), “Do Reward Programs Build Loyalty for Services? The Moderating Effect of Satisfaction on Type and Timing of Rewards”, Journal of Retailing, 82 (June), 127-136. Kopalle, Praveen K. and Scott A. Neslin 2003, “The Economic Viability of Frequency Reward Programs in a Strategic Competitive Environment”, Review of Marketing Science, 1, 1-39. Leenheer, Jorna, Harald J. van Heerde, Tammo H.A. Bijmolt, and Ale Smidts (2007), “Do Loyalty Programs Really Enhance Behavioral Loyalty? An Empirical Analysis Accounting for Self-selecting Members”, International Journal of Research in Marketing, 24 (March), 31-45. Liu, Yuping (2007), “The Long-Term Impact of Loyalty Programs on Consumer Purchase Behavior and Loyalty”, Journal of Marketing, 71 (October), 19–35. Mauri, Chiara (2003), “Card loyalty. A New Emerging Issue in Grocery Retailing”, Journal of Retailing and Consumer Services, 10 (January), 13–25. McConnell, J. Douglas (1968), “The Development of Brand Loyalty: An Empirical Study”, Journal of Marketing Research, 5 (February), 13 - 19. Meyer-Waarden, Lars (2007), “The Effect of loyalty Programs on Customer Lifetime duration and share of wallet”, Journal of Retailing, 83 (April), 223-236. Nord, Walter R. and Peter J. Paul (1980), A Behavior Modification Perspective On Marketing, Journal of Marketing, 44, (Spring 1980), 36-47. O’Brien, Louise and Jones Charles (1995), “Do Rewards Really Create Loyalty”, Harvard Business Review, May-June, 75-82. Odin, Yorrick. (1998), Fidélité et Inertie: Clarification Conceptuelle et Test Empirique, Doctoral Thesis in Management Sciences. Oliver, Richard L. (1999), “Whence Consumer Loyalty ?” Journal of Marketing, 63 (October), 33-44.

48 ISSN 1961-7798 - © 2009, International Marketing Trends Conference

Reinares, Lara Pedro and Jesús García de Madariaga, (2007), “The importance of rewards in the management of multisponsor loyalty programmes”, Journal of Database Marketing & Customer Strategy Management, 15, 37–48 Roehm, Michelle L, Ellen Bolman Pullins and Roehm Harper A. (2002), “Doing Loyalty-Building Programs for Packaged Goods Brands”, Journal of Marketing Research, 39 (May), 202- 213. Rothschild, Michael L. and William C. Gaidis (1981), “Behavioral Learning. Theory: its Relevance to Marketing and Promotions”, Journal of Marketing, 45 (Spring) 70-78. Sharp, Byron and Anne Sharp (1997), “Loyalty Programs and Their Impact on Repeat-Purchase Loyalty Patterns”, International Journal of Research in Marketing, 14 (December), 473-486. Shugan, Steven M. (2005), “Brand Loyalty Programs: Are They Shams?” Marketing Science, 24 (Spring), 185-193. Skinner, Burrhus F. (1965), Science and Human Behavior, New York, Free Press. Stafford James E. (1966), Effect of Group Influences on Consumer Brand Preference, Journal of Marketing Research, 3 (February), 68 - 74. Taylor, Gail Ayala and Scott A. Neslin (2005), “The current and future sales impact of a retail frequency reward program,” Journal of Retailing, 81, 4, 293-305. Trinquecoste, Jean-François (1996), Fidéliser le consommateur: un objectif marketing prioritaire, Décisions Marketing, 7, January-April, 17-23. Tucker W.T (1964) “The Development of Brand Loyalty”, Journal of Marketing Research, (August), 32-35. Verhoef, Peter C. (2003), “Understanding the Effect of Customer Relationship Management Efforts on Customer Retention and Customer Share Development”, Journal of Marketing, 67 (October), 30–45. Yi, Youjae and Hoseong Jeon (2003), “Effects of Loyalty Programs on Value Perception, Program Loyalty and Brand Loyalty”, Journal of the Academy of Marketing Science, 31 (Summer), 229-240.


24366_int_Mise en page 1 22/12/10 12:25 Page49

JOURNAL OF MARKETING TRENDS - TOURISM

Determinantes y consecuentes de la satisfacción hacia el turismo rural: el efecto moderador de la experiencia previa Determinants and consequences of satisfaction with rural tourism: the moderating effect of previous tourists’ experience En este trabajo se pretende analizar los determinantes y consecuentes de la satisfacción hacia el turismo rural mediante la conformación de un modelo global, contemplando el efecto moderador que tiene la experiencia previa del turista en la conformación de la satisfacción con el turismo rural realizado. Se ha planteado un modelo integrado de los determinantes y consecuentes de la satisfacción global usando indicadores formativos y reflectivos mediante el análisis PLS. Para cumplir con este objetivo se ha llevado a cabo un estudio empírico sobre una muestra de 164 turistas rurales. Los resultados muestran que para los turistas que han visitado el destino turístico rural en otras ocasiones la satisfacción con las fuentes de información no tiene ninguna influencia sobre la satisfacción global hacia el destino. Estos resultados proporcionan apoyo empírico para la literatura previa que ha hallado como son las fuentes de información internas las principales determinantes del comportamiento de revisita al destino turístico. The aim of this study is to analyze the determinants and consequences of satisfaction with rural tourism by designing a global model that considers the moderating effect of tourists’ previous experience with the chosen tourist destination. We have carried out an integrated model of the determinants and consequences of overall satisfaction using formative and reflective indicators together by means of the PLS method. To this aim an empirical study was carried out using a sample of 164 customers of rural tourism. The results shows that for individuals that have visited the rural tourist destination on other occasions, satisfaction with information sources has no influence on overall satisfaction. These results provide empiric support for the literature describing internal information sources as the main resource for tourists repeating their holiday destination. Palabras clave: Turismo, turismo rural, fuentes de información, satisfacción, valor percibido. Key words: multi-sponsor loyalty program, latent initial purchase tendency, latent tendency to control the initial purchase. > Salvador Del Barrio García University of Granada, Spain dbarrio@ugr.es

> Dolores María Frías Jamilena University of Granada, Spain dfrias@ugr.es

> Lorenza López Moreno University of Granada, Spain llopezm@ugr.es Journal of Marketing Trends - Volume I (February 2011) 49


24366_int_Mise en page 1 22/12/10 12:25 Page50

JOURNAL OF MARKETING TRENDS - TOURISM

INTRODUCCIÓN

Introduction

En los últimos años, el desarrollo del turismo rural se ha visto favorecido por un mayor dinamismo de la demanda, motivado en gran parte por el mayor peso que actualmente presentan las motivaciones vinculadas al interés por la conservación del medio ambiente y el contacto directo con la naturaleza (Hall 2004), así como el avance de la sociedad hacia ciertos valores como la personalización y la tranquilidad. Sin embargo, este desarrollo favorable de la demanda debe ser analizado con atención ya que cada vez existe una mayor oferta turística, tanto cuantitativa como cualitativa, habiendo aumentado la variedad y el número de destinos alternativos lo que hace difícil mantener la cuota de mercado para los diferentes destinos turísticos. La competitividad de los diferentes destinos turísticos hace que sea clave examinar las variables que influyen en la satisfacción con el destino turístico y en las intenciones comportamentales de los turistas. Numerosas investigaciones han demostrado que la información es un valioso concepto en el entendimiento del proceso de elección de un destino. Algunos estudios se han centrado en la relación entre las fuentes de información y la selección actual del destino (Um y Crompton 1990). Otros trabajos han investigado las de información en el proceso de decisión del destino (Baloglu 2000; Capella y Greco 1987; Sirakaya y Woodside 2005), o la influencia de dichas fuentes en las preferencias e intenciones de los turistas (Baloglu 2000; Um y Crompton 1999). En cambio, son escasos los trabajos que se han centrado en analizar la influencia que la adecuación o satisfacción con las fuentes de información utilizadas para seleccionar un destino ejerce sobre la satisfacción con el destino turístico, en ningún caso considerando de forma desagregada las distintas fuentes que el turista puede consultar antes de visitar el destino turístico ni referido al turismo rural. El valor percibido adquiere una gran importancia en la gestión de marketing, puesto que los directivos deben conocer lo que valora el cliente y en qué centra su atención, llegando a ser destacado como uno de los antecedentes de la satisfacción (González y Gil 2006; González, Arteaga y Gil 2005). Identificándose la satisfacción como el requisito fundamental para que exista lealtad, tanto en la literatura general de marketing como en la relacionada de forma específica con el ocio y el turismo, lo que justifica la importancia de su estudio, y orienta los objetivos de la presente investigación. No obstante, en la literatura se muestra que existen diferencias en cuanto al proceso de decisión de compra entre los individuos que compran un producto o servicio por primera vez de aquellos que lo han hecho en otras ocasiones (Alba y Hutchinson 1987), o cuando se trata de individuos que visitan un destino por primera vez y cuando repiten su visita (Spence y Bruces 1997).

In recent years, the development of rural tourism has been encouraged by more dynamic demand, to a large extent caused by higher interest in environmental conservation and direct contact with nature (Hall 2004), as well as a move in society towards values such as personalization and tranquillity. However, this favourable development of demand should be analysed carefully as there is an increasingly large tourist offer in both quantitative and qualitative terms, with an increase in the variety and number of alternative destinations, making it hard for the different tourist destinations to maintain their market share. The competitivity of the different tourist destinations makes it imperative to examine the variables affecting satisfaction with the destination and the tourists’ behavioural intentions. Many studies have shown that information is a valuable concept for understanding the process of choosing a destination. Some of these have focussed on the relation between the information sources and the actual choice of destination (Um & Crompton 1990). Other studies have researched information sources in the process of destination choice (Baloglu 2000; Capella & Greco 1987; Sirakaya & Woodside 2005), or the influence of such sources on tourists’ preferences and intentions (Baloglu, 2000; Um & Crompton, 1999). However, few studies have centred on analyzing the influence that suitability or satisfaction with the information sources used to select a destination has on satisfaction with the tourist destination, and none have considered in isolation the different sources the tourist may have consulted before visiting their chosen destination, nor has the focus been on rural tourism. Perceived value takes on great importance in marketing management, and becomes an outstanding antecedent of satisfaction (González & Gil 2006; González, Arteaga & Gil 2005). Satisfaction can be identified as the basic requirement for loyalty, both in the general literature on marketing and in literature specifically concerned with leisure and tourism, thus justifying the importance of its study and orienting the aims of the present research. Both sources of information and perceived value are concepts influencing choice of destination, although the literature shows that there are differences between consumers purchasing a product or service for the first time and those who have done so on other occasions (Alba & Hutchinson 1987), as too between individuals visiting a destination for the first time and those making a repeat visit (Spence & Brucks 1997). However, although studies have shown these differences to exist, they have not clearly identified their causes. This research therefore has a double objective. First, it attempts to analyze the extent to which the tourist’s

50 ISSN 1961-7798 - © 2009, International Marketing Trends Conference


24366_int_Mise en page 1 22/12/10 12:25 Page51

JOURNAL OF MARKETING TRENDS - TOURISM

Así pues, este trabajo presenta un doble objetivo. En primer lugar, se trata de analizar en qué medida afecta la satisfacción que ha tenido el turista con las distintas fuentes de información consultadas y el valor percibido en la satisfacción global con la actividad turística rural desarrollada y, por ende, en sus intenciones comportamentales. Un segundo objetivo tiene que ver con el análisis del efecto moderador que tiene la experiencia previa del turista en la conformación de la satisfacción con la actividad de turismo rural.

REVISIÓN DE LA LITERATURA Los turistas utilizan la información externa en una primera etapa cuando planean su viaje, de ahí que sea muy importante que profesionales y promotores tengan información acerca de las necesidades que demanda cada turista de cada una de esas fuentes para intentar mejorar su contenido y aumentar su eficacia (Molina y Águeda 2006). Para este colectivo es importante entender cómo la información que proporcionan al turista puede influir en la satisfacción global con el destino. Según algunos autores un antecedente de la satisfacción global es la satisfacción con la información (Petrick y Backman 2002; Spreng, MacKenzie y Olshavsky 1996). Si bien es cierto que hay otros antecedentes de la satisfacción éste es el único que permite conocer la evaluación que hace el consumidor de los esfuerzos de comunicación realizados por la empresa para atraer a los turistas a un destino turístico. La relación entre satisfacción con la información y satisfacción global ha sido propuesta y probada en diferentes contextos: satisfacción del consumidor con la web (McKinney, Yoon y Zahedi 2002), e-satisfaction (Szymanski y Hise 2000), sistemas de información (DeLone y McLean 1992), productos tangibles (Spreng et al. 1996), o turismo (Castañeda, Frías y Rodríguez 2007; Petrick y Beckman 2002). Una de las principales conclusiones de los trabajos aplicados a turismo es la importancia de la satisfacción con la información por parte del turista como condición para tener una experiencia gratificante y finalmente repetir la compra del servicio. En este sentido la calidad y cantidad de información obtenida durante el proceso de decisión debería tener un impacto positivo sobre la satisfacción con el destino (Peterson, Balasubramanian y Bronenberg 1997; Szymanski y Hise 2000). Por tanto, sobre la base de la discusión anterior cabe plantear las siguientes hipótesis: H1a: Cuanto mayor sea la satisfacción con la información obtenida por el turista a través de fuentes comerciales impersonales (publicidad) mayor será su satisfacción con el destino seleccionado para hacer turismo rural. H1b: Cuanto mayor sea la satisfacción con la información obtenida por el turista a través de fuentes no comerciales impersonales (publicidad no pagada)

satisfaction with the various information sources consulted affects overall satisfaction with the rural tourism activity carried out and, therefore, the tourist’s behavioural intentions, as well as the effect of perceived value on their satisfaction and behavioural intentions. The second aim is concerned with the analysis of the moderating effect of the tourist’s previous experience on shaping his satisfaction with the rural tourism activity.

REVIEW OF THE LITERATURE Tourists use external information in the initial stage of planning their journey, so it is very important for professionals and promoters to have information on the needs of each tourist as regards each of these sources in order to try to improve their content and increase their efficiency (Molina & Águeda 2006). For this collective it is important to understand how the information provided to the tourist can influence his overall satisfaction with the destination. Some authors consider that satisfaction with information is an antecedent to global satisfaction (Petrick & Backman, 2002; Spreng, MacKenzie & Olshavsky, 1996). Although it is true that there are other antecedents of satisfaction, this is the only one that allows us to know the consumer’s evaluation of the communication efforts made by the enterprise to attract tourists to a particular destination. The relation between satisfaction with the information and overall satisfaction was proposed and tested in different contexts: consumer satisfaction with the web (McKinney, Yoon & Zahedi 2002), e-satisfaction (Szymanski & Hise 2000), information systems (DeLone & McLean 1992), tangible products (Spreng et al. 1996) and tourism (Castañeda, Frías & Rodríguez 2007; Petrick & Beckman 2002). One of the main conclusion of the studies on tourism is the importance of tourist satisfaction with information as a condition for having a gratifying experience and, finally, repeating the purchase of the service. In this sense, the quality and quantity of information obtained during the decision process should have a positive impact on satisfaction with the destination (Peterson, Balasubramanian & Bronenberg 1997; Szymanski & Hise 2000). Therefore, on the basis of the foregoing discussion, we can put forward the following hypotheses: H1a: The greater the satisfaction obtained by the tourist through impersonal commercial sources (publicity), the greater will be his satisfaction with the destination chosen for rural tourism. H1b: The greater the satisfaction obtained by the tourist through impersonal non-commercial sources (unpaid publicity), the greater will be his satisfaction with the destination chosen for rural tourism. H1c: The greater the satisfaction obtained by the

Journal of Marketing Trends - Volume I (February 2011) 51


24366_int_Mise en page 1 22/12/10 12:25 Page52

JOURNAL OF MARKETING TRENDS - TOURISM

mayor será su satisfacción con el destino seleccionado para hacer turismo rural. H1c: Cuanto mayor sea la satisfacción con la información obtenida por el turista a través de fuentes comerciales personales (agencias de viaje, oficinas de turismo y ferias o eventos de turismo), mayor será su satisfacción con el destino seleccionado para hacer turismo rural. H1d: Cuanto mayor sea la satisfacción con la información obtenida por el turista a través de fuentes no comerciales personales (amigos y familiares), mayor será su satisfacción con el destino seleccionado para hacer turismo rural. H1e: Cuanto mayor sea la satisfacción con la información obtenida por el turista a través de Internet mayor será su satisfacción con el destino seleccionado para hacer turismo rural. El valor percibido ha sido reconocido como uno de los principales antecedentes de la satisfacción (González et al. 2005; González y Gil 2006). En este sentido, Oh (1999) desarrolló un modelo integrado y concluyó que el valor percibido es un antecedente importante de la satisfacción. También se ha analizado la relación entre el valor percibido y la intención de compra (Oh 1999; Petrick y Backman 2002). Wakefield y Barnes (1996) encuentran que la calidad percibida en el servicio influye en el valor percibido del servicio, y que el valor percibido tiene una influencia positiva en la intención de repetir. De este modo, sugieren que la mejora en el entorno de los servicios puede incrementar la percepción de calidad del cliente, lo que puede mejorar su patrón de repetición. De acuerdo a estos hallazgos proponemos las siguientes hipótesis: H2: El valor percibido por el turista de un alojamiento rural tendrá una influencia positiva en su satisfacción con el alojamiento turístico. H3: El valor percibido por el turista de un alojamiento rural tendrá una influencia positiva en su grado de fidelidad o intención de comportamiento hacia el alojamiento turístico. La influencia positiva de la satisfacción sobre las intenciones ha sido ampliamente sugerida y probada tanto en la literatura del marketing (Petrick 2004; Tian-Cole, Crompton y Willson 2002), en general, como en turismo (Cronin y Taylor 1992; Fornell 1992; Oliver y Swan 1989), en particular. En este sentido, autores como Bigné, Sánchez y Sánchez (2001), Lee et al. (2005) y Chen y Gursoy (2007), hallaron que la satisfacción influye sobre la recomendación del destino a terceros, de tal forma que individuos con alta satisfacción en su experiencia de viaje muestran unas intenciones de comportamiento más positivas. A la vista de lo anterior se plantea una nueva hipótesis: H4: Cuanto mayor sea la satisfacción con el destino de turismo rural mayor será la fidelidad o intención de comportamiento hacia el mismo.

52 ISSN 1961-7798 - © 2009, International Marketing Trends Conference

tourist through personal commercial sources (travel agencies, tourist offices and tourism fairs or events), the greater will be his satisfaction with the destination chosen for rural tourism. H1d: The greater the satisfaction obtained by the tourist through personal non-commercial sources (friends and relatives), the greater will be his satisfaction with the destination chosen for rural tourism. H1e: The greater the satisfaction obtained by the tourist through Internet, the greater will be his satisfaction with the destination chosen for rural tourism. Perceived value has been recognized as one of the main antecedents of satisfaction (González et al. 2005; González & Gil 2006; Woodruff 1997). In this sense, Oh (1999) developed an integrated model, concluding that perceived value is an antecedent of satisfaction. The relation between perceived value and purchase intention has also been analyzed (Oh 1999; Petrick & Backman 2002). Wakefield & Barnes (1996) found that the perceived quality of the service influences the perceived value of the service, and that the perceived value has a positive effect on the intention to repurchase. They therefore suggested that the improvement in the field of services could increase the client’s perception of quality, which in turn could improve his repetition pattern. In this sense, we propose the following hypotheses: H2: The value perceived by a tourist in a rural accommodation will have a positive influence on his satisfaction with that accommodation. H3: The value perceived by a tourist in a rural accommodation will have a positive influence on the degree of loyalty or intention of behaviour towards that accommodation. The positive influence of satisfaction on intentions has been widely suggested and proved in both general marketing literature (Petrick 2004; Tian-Cole, Crompton & Willson 2002) and literature specifically on tourism (Cronin & Taylor 1992; Fornell 1992; Oliver & Swan 1989). Authors such as Bigné, Sánchez & Sánchez (2001), Chen & Tsai (2007) and Lee, Lee & Lee (2005) found that satisfaction influences the recommendation of a destination to third parties, so that individuals with high satisfaction of their travel experience show more positive intentions of behaviour. In view of the foregoing, we can suggest another hypothesis: H4: The higher the satisfaction with a rural tourism destination, the higher will be the loyalty or intention of behaviour towards that destination. Many studies have shown that there are differences between expert and novice consumers in the process of deciding to purchase (Alba & Hutchinson 1987; Bettman & Sujan 1987; Groen & Patel 1988; Johnson 1988; Rao & Sieben 1992; Spence & Brucks 1997). The information sources used by consumers depend


24366_int_Mise en page 1 22/12/10 12:25 Page53

JOURNAL OF MARKETING TRENDS - TOURISM

Numerosas investigaciones han demostrado que existen diferencias en cuanto al proceso de decisión de compra entre consumidores expertos y noveles (Alba y Hutchinson 1987; Bettman y Sujan 1987; Groen y Patel 1988; Johnson 1988; Rao y Sieben 1992; Spence y Brucks 1997). Las fuentes de información utilizadas por los consumidores para buscar información dependen de las características de los productos y servicios que se estén comercializando. El turista para planear su viaje o repetir la visita a un destino acude a las fuentes de información internas, como su experiencia pasada con el destino específico o con otro destino similar (Chen y Gursoy 2001; Schul y Crompton 1983), de tal manera que si la información disponible en su memoria no es suficiente para tomar la decisión acude a fuentes de información externas (Beatty y Smith 1987; Schul y Crompton 1983). Esto supone que para el turista con una experiencia previa en el destino el efecto de la satisfacción con las fuentes de información sobre la satisfacción global será menor que en el caso de individuos que no tienen experiencia previa y tienen que hacer una búsqueda externa (Schul y Crompton 1983). De ello se deriva la sexta hipótesis de nuestra investigación: H5: La experiencia previa con un destino modera el efecto de la satisfacción con las fuentes de información consultadas sobre la satisfacción del destino turístico. Del mismo modo, la experiencia previa con el destino debe afectar a la relación entre el valor percibido, la satisfacción y la intención de comportamiento. Petrick (2004), aplicado al turismo de cruceros, trata de identificar las diferencias entre los visitantes que acuden por primera vez y los que repiten. Los resultados de su estudio revelaron la existencia de diferencias significativas para los antecedentes del valor percibido entre los visitantes noveles y los más expertos, así como que la calidad era el mejor predictor de la intención de recompra para el caso de los pasajeros que realizaban el crucero por primera vez, mientras que el valor percibido lo era para los que repetían. En la revisión de la literatura se pone de manifiesto que tanto la persuasión que hay que realizar para que el turista visite un alojamiento, como su satisfacción e intención comportamental se van a ver condicionados por su experiencia previa. En este sentido, proponemos contrastar las siguientes hipótesis: H6: La visita previa de un destino modera el efecto del valor percibido sobre la satisfacción. H7: La visita previa de un destino modera el efecto del valor percibido sobre la intención de comportamiento. H8: La visita previa de un destino modera el efecto directo de la satisfacción sobre la intención de comportamiento.

on the characteristics of the products and services commercialised. In order to plan a trip or repeat a visit to a particular destination, the tourist turns to internal information sources, such as his past experience with that destination or another similar one (Chen & Gursoy 2001; Schul & Crompton 1983), so that if the information available in his memory is not sufficient to take the decision, he will turn to external sources (Beatty & Smith 1987; Schul & Crompton 1983). This means that for the tourist with prior experience of the destination, the effect of satisfaction with information sources on his overall satisfaction will be less than in the case of individual with no prior experience, who must make an external search (Schul & Crompton 1983). We can thus propose the sixth hypothesis of this research: H5: Prior experience with a destination moderates the effect of satisfaction with the information sources consulted on satisfaction with the destination in question. Likewise, prior experience with the destination should affect the relation between perceived value, satisfaction and intention of behaviour. Considering cruising tourism, Petrick (2004) attempted to identify the differences between first-time travellers and those repeating. The results showed significant differences for the antecedents of perceived values between novice and more expert travellers, and also that the best predictor of intention to repurchase was quality in the case of first-time passengers, while it was perceived value for those repeating. In the review of the literature, it is shown that both the persuasion required to make the tourist visit an accommodation, and his satisfaction and behavioural intention are conditioned by his prior experience. In accordance with the foregoing, we propose the following hypotheses: H6: A prior visit to a destination moderates the effect of the perceived value on satisfaction. H7: A prior visit to a destination moderates the effect of the perceived value on behavioural intention. H8: A prior visit to a destination moderates the direct effect of satisfaction on behavioural intention.

METHOD AND EMPIRICAL SETTING Geographical location of the study: Andalusia The geographical location for this research is the Autonomous Community of Andalusia, as it is one of the main tourist destinations in Spain. Specifically, in 2006 Andalusia received a total of 36.4 million foreign and national tourists, representing 17% of the national total (second destination chosen after Catalonia), involving 16.2% of all the overnight stays consumed in Spain with an average length of 5.9 nights (0.3 nights above the national average) (Institute of Tourist Studies, IET 2007). On the national level, the rural and nature tourism segment received over 2.6 million tourists in 2007 (90%

Journal of Marketing Trends - Volume I (February 2011) 53


24366_int_Mise en page 1 22/12/10 12:25 Page54

JOURNAL OF MARKETING TRENDS - TOURISM

METODOLOGÍA Y ESTUDIO EMPÍRICO Ámbito geográfico del estudio: Andalucía El ámbito geográfico utilizado en esta investigación es la Comunidad Autónoma de Andalucía por ser el principal destino elegido por los españoles. Concretamente, en 2006 Andalucía recibió un total de 36,4 millones de turistas extranjeros y nacionales, representando el 17% del total nacional y suponiendo el 16,2% del número de pernoctaciones consumidas en España con una estancia media que se sitúa en 5,9 noches (IET 2007). A nivel nacional, el segmento de turismo rural y de la naturaleza alcanzó los 2,6 millones de turistas en 2007 (90% españoles, 10% extranjeros), con caso 8 millones de pernoctaciones. Con un total de 171.734 turistas, Andalucía era el 4º destino nacional más popular detrás de Castilla y León, Cataluña y Cantabria.

Metodología Para cumplir con los objetivos propuestos en la investigación se diseñó una metodología basada en la recogida de información mediante cuestionarios cerrados y estructurados a una muestra de turistas que habían realizado al menos una estancia en alojamientos turísticos rurales de Andalucía. Para ello se contó con la colaboración de la Asociación de Hoteles Rurales de Andalucía (AHRA) que cuenta entre sus filas con más de 60 establecimientos rurales repartidos por toda la geografía andaluza. AHRA proporcionó un listado de 243 clientes con predisposición a participar en una investigación sobre turismo rural. Transcurridas unas semanas se contactaba con los turistas vía telefónica y se les pasaba el cuestionario. Finalmente, se consiguieron 164 cuestionarios válidos que conforman la muestra final. La tabla 1 muestra la ficha técnica del estudio. TABLA 1 - Ficha técnica

Cuestionario y escalas de medida A partir de la revisión de la literatura se confeccionó un cuestionario que recogía los conceptos objeto de estudio, medidos a través de diversos ítems que componían escalas de diferencial semántico de 5 puntos

54 ISSN 1961-7798 - © 2009, International Marketing Trends Conference

Spanish nationals, 10% foreigners) with almost 8 million overnight stays. With a total of 171,734 tourists (6.45%) Andalusia was the 4th most popular destination after Castille and Leon, Catalonia and Cantabria.

Methodology To achieve the aims proposed in this research, a methodology was designed, based on information collection using closed, structured questionnaires for a sample of tourists that had undertaken at least one stay in rural tourist accommodations in Andalusia. For this purpose we had the collaboration of the Association of Rural Hotels in Andalusia (AHRA), with over 60 affiliated rural establishments throughout Andalusia. AHRA thus provided a list of 243 clients willing to take part in research on rural tourism. A few weeks later, the tourists were contacted by phone and sent the questionnaire. As a result, 164 valid questionnaires were obtained making up the final sample. Table 1 shows the technical specifications of the study. TABLE 1 - Technical Data

Questionnaire and measurement scales A questionnaire was designed based on the literature review, containing the concepts to be studied and measured through various items making up differential semantic scales of 5 points in all cases. To measure satisfaction with impersonal commercial information sources a single item was used: [Degree of satisfaction with mass media publicity (television, radios, press…) (FICIMP1]. Likewise, a single item was used to measure satisfaction with impersonal non-commercial information sources, [Degree of satisfaction with nonadvertising information appearing in the media (news in press, television, radio, etc.) (FINCIMP1)], and with personal non-commercial information sources, [Degree of satisfaction with information provided by friends and relatives (word-of-mouth) (FINCPERS1)]. Measurement of the rural tourist’s satisfaction with personal commercial and web information sources the following multi-item scales were used. In the first case: Degree of satisfaction with … [information provided by


24366_int_Mise en page 1 22/12/10 12:25 Page55

JOURNAL OF MARKETING TRENDS - TOURISM

en todos los casos. Para medir la satisfacción hacia las fuentes de información comerciales impersonales se utilizó un único ítem: [Grado de satisfacción hacia la publicidad en medios masivos (televisión, radio, prensa…) (FICIMP1)]. Para medir la satisfacción con las fuentes de información no comerciales impersonales: [Grado de satisfacción hacia la información no publicitaria que aparece en los medios de comunicación (noticias en prensa, televisión, radio…) (FINCIMP1)], y con las fuentes de información no comerciales personales: [Grado de satisfacción hacia la información proporcionada por amigos y familiares (boca-boca) (FINCPERS1)]. Para medir la satisfacción del turista rural hacia las fuentes de información personales comerciales y web se han empleado las siguientes escalas multi-ítem. En el primer caso: Grado de satisfacción hacia la…[ información proporcionada por el personal de las agencias de viajes (FICPERS1)], [información proporcionada por el personal de las oficinas de turismo (FICPERS2)], [información proporcionada en ferias o eventos de turismo (FICPERS3)]. En el segundo caso: Grado de satisfacción hacia…[sitios web de las agencias de viajes (FIWEB1)], [sitios web de los propios establecimientos turísticos (FIWEB2)], sitios web institucionales (Ministerio, Consejerías, etc.) (FIWEB3)], [los foros de Internet (FIWEB4)]. Para medir el valor percibido se utilizó una escala tipo Likert de dos ítems, basada en la propuesta por Cronin et al. (2000), y aplicada en diferentes trabajos desarrollados en el ámbito turístico (González et al. 2005; González y Gil 2006): Evalúe el grado de acuerdo o desacuerdo con las siguientes afirmaciones de su últimas vacaciones rurales en Andalucía (1: totalmente en desacuerdo – 5: totalmente de acuerdo): [Comparando los sacrificios hechos con los beneficios obtenidos, considero que la experiencia turística ha sido adecuada (VP1)], [Comparado con lo que he tenido que sacrificar, esta experiencia turística me ha permitido satisfacer mis deseos y necesidades de forma adecuada (VP2)]. La satisfacción global con el destino turístico fue medida mediante una escala de tres ítem de diferencial semántico de 5 puntos adaptada de Spreng et al. (1996) y aplicada en la investigación en turismo por Baker y Crompton (2000): ¿Cómo evaluaría sus últimas vacaciones de turismo rural en Andalucía?... [Desagradables – Agradables (ST1)], [Negativas – Positivas (ST2)], [Insatisfactorias – Satisfactorias (ST3)]. Las intenciones comportamentales con el destino de turismo rural se midieron con una escala tipo Likert de dos ítem ampliamente utilizados en la literatura de turismo (Bigné et al. 2001; Tian-Cole et al. 2002): Evalúe su grado de acuerdo o desacuerdo con las siguientes afirmaciones (1: totalmente en desacuerdo- 5: totalmente de acuerdo): [Recomendaré este tipo de viaje a familiares y amigos (ITC1)], [Volveré a hacer este tipo de viaje en un futuro (ITC2)].

travel agency staff (FICPERS1)], [information provided by tourist office staff (FICPERS2)], [information provided at tourism fairs or events (FICPERS3)]. In the second case: Degree of satisfaction with ... [travel agency web sites (FIWEB1)], [the tourist establishments’ own web sites (FIWEB2)], institutional web sites (Ministry or Departments of Tourism, etc.)(FIWEB3)], [Internet forums (FIWEB4)]. A two-item Likert-type scale was used to measure perceived value, based on the proposal by Cronin et al. (2000), and applied in various studies of the tourist sector (González & Gil 2006; González et al. 2005): Evaluate your degree of agreement or disagreement with the following statements about your last rural holidays in Andalusia (1: totally disagree – 5: totally agree): [Comparing the sacrifices made and the benefits obtained, I consider the tourist experience to have been adequate (VP1)], [Compared with what I had to sacrifice, this tourist experience allowed me to adequately satisfy my desires and needs (VP2)]. Overall satisfaction with the tourist destination was measured using a three-item 5-point semantic differential scale adapted from Spreng et al. (1996), applied in research on tourism by Baker and Crompton (2000): How would you evaluate your last rural tourism holiday in Andalusia? … [Unpleasant – Pleasant (ST1)], [Negative – Positive (ST2)], [Unsatisfactory – Satisfactory (ST3)]. Behavioural intentions regarding the rural tourism destination were measured using a two-item Likert-type scale widely used in the literature (Bigné et al. 2001; Sirakaya et al. 2001; Tian-Cole et al., 2002): Evaluate your degree of agreement or disagreement with the following statements (1: complete disagreement – 5: complete agreement): [I shall recommend this type of trip to relatives and friends (ITC1)], [I shall make another trip of this sort in the future (ITC2)]. The last part of the questionnaire contains different variables of classification, such as sex, age, level of education, income level, place of origin and degree of experience with the tourist destination visited. The aim of this last question was to divide the sample into two groups according to whether this was their first experience of visiting the rural tourism destination (novices) or not (experts).

Theoretical model of determinants and consequences of satisfaction with rural tourism In order to fulfil the proposed objectives and empirically compare the research hypotheses formulated above, a causal model was designed around the concept of satisfaction with the rural tourism activity carried out (see Figure 1). According to hypotheses H1a, H1b, H1c, H1d and H1e, rural tourists’ satisfaction with the external information sources consulted will affect their overall satisfaction with the tourist activity undertaken. For this

Journal of Marketing Trends - Volume I (February 2011) 55


24366_int_Mise en page 1 22/12/10 12:25 Page56

JOURNAL OF MARKETING TRENDS - TOURISM

La parte final del cuestionario recogía diversas variables de clasificación como el sexo, la edad, el nivel de estudios, el nivel de ingresos, el lugar de procedencia y el nivel de experiencia con el destino turístico visitado. Con esta última cuestión se pretendía dividir a la muestra en dos grupos en función a si era la primera vez que visitaban el destino de turismo rural (noveles) o no (expertos).

Modelo teórico de determinantes y consecuentes de la satisfacción hacia el turismo rural Al objeto de dar respuesta a los objetivos propuestos y poder contrastar de manera empírica las hipótesis de investigación formuladas en apartados anteriores se desarrolló un modelo causal en torno al concepto de satisfacción con la actividad de turismo rural realizada (Véase figura 1).

study the external information sources were divided into five groups, depending on whether they were commercial or not and whether personal or not (Fodness & Murray 1997). The distinction was therefore made between information sources that were impersonal and commercial (FICIMP), impersonal and non-commercial (FINCIMP), personal and commercial (FICPER), and personal and non-commercial (FNCPER). A fifth group was added to these four types, corresponding to information sources obtained from web sites (FIWEB), such as travel agencies, tourist establishments, institutions and Internet forums, which, because of their increasing importance in the sector, are worth analyzing separately from the others. FIGURE 1 - Theoretical model of determinants and consequences of satisfaction with rural tourism

FIGURA 1 - Modelo de antecedentes y consecuentes de la satisfacción hacia el turismo rural

De acuerdo a las hipótesis H1a, H1b, H1c, H1d, H1e la satisfacción que tengan los turistas rurales con las fuentes de información externas consultadas influirá en la satisfacción global con la actividad turística desarrollada. En nuestra investigación se han dividido las fuentes de información externas en cinco grupos, atendiendo a si las fuentes tienen o no carácter comercial y si son o no personales (Fodnees y Murray 1997). Así se ha diferenciado entre fuentes de información comerciales impersonales (FICIMP), fuentes de información no comerciales impersonales (FINCIMP), fuentes de información comerciales personales (FICPER), y fuentes de información no comerciales personales (FNCPER). A estos cuatro tipos de fuentes de información externas se les ha añadido un quinto grupo que son las fuentes de información obtenidas de los sitios web (FIWEB) tales como agencias de viajes, establecimientos turísticos, instituciones y foros de Internet, que por su importancia creciente en este sector merecen ser analizadas de forma desagregada de las anteriores. Otro antecedente significativo de la satisfacción

56 ISSN 1961-7798 - © 2009, International Marketing Trends Conference

Another significant antecedent of overall satisfaction according to H2 is the value perceived by the tourist of the activity undertaken (VALORPERC). Finally, and to complete the causal model, we established the relations between overall satisfaction with the rural tourism destination and behavioural intentionality (H4) and between perceived value and overall satisfaction (H2) and behavioural intention (H3). To estimate the proposed model we use the Structural Equations Modelling by the PLS technique. The causal model proposed in this study is therefore much better adapted to the conditions for application of the Partial Least Squares technique (PLS) than the Methods based on Covariance Analysis (MBC).

RESULTS For constructs with reflective indicators, their attributes of validity and reliability were evaluated for each of the two groups (experts against novices). Table 2 shows the adequate internal consistency of the scales used to measure perceived value, satisfaction and behavioural intention in both groups. All the charges were significant and the compound reliability and extracted variance values were above the recommended limits (0.8 and 0.5, respectively).


24366_int_Mise en page 1 22/12/10 12:25 Page57

JOURNAL OF MARKETING TRENDS - TOURISM

global de acuerdo a H2 es el valor que percibe el turista de la actividad desempeñada (VALORPERC). Finalmente, y para completar el modelo causal se han establecido las relaciones entre la satisfacción global con el destino de turismo rural y la intencionalidad comportamental (H4), del valor percibido sobre la satisfacción (H2) y del valor percibido sobre la intención comportamental (H3). Para estimar el modelo propuesto se empleó el análisis de ecuaciones estructurales mediante la técnica estadística PLS (Partial Least Squares) ya que nuestra investigación se adapta muy bien a las condiciones de aplicación de la técnica PLS por diversos motivos.

RESULTADOS Para los constructos con indicadores reflectivos se procedió a evaluar sus atributos de validez y fiabilidad para cada uno de los dos grupos (expertos versus noveles). La tabla 2 pone de manifiesto la adecuada consistencia interna de las escalas utilizadas para medir el valor percibido, la satisfacción y la intención de comportamiento en ambos grupos. Todas las cargas resultaron significativas y los valores de fiabilidad compuesta y varianza extraída estaban por encima de los límites recomendados (0,8 y 0,5, respectivamente). Asimismo, se comprobó la validez discriminante de los tres constructos en ambos grupos de acuerdo al procedimiento propuesto por Fornell y Larcker (1981) según el cual la raíz cuadrada de las varianzas extraídas debe ser mayor a las correlaciones entre constructos. TABLA 2 - Atributos del modelo de medida con indicadores reflectivos

TABLE 2 - Measurement Model Attributes with Reflective Indicators

Likewise, the discriminant validity of the three constructs in each group was tested following the procedure proposed by Formell & Larcker (1981), by which the square root of the extracted variances must be greater than the correlations between constructs. For the formative measurements is important to evaluate the possibility of multicolinearity of the formative indicators, given that high colinearity between them would lead to problems for separating the different influence of each individual indicator on the latent variable. With this aim we calculated the inflation factor of the variance (FIV) for each group and the tolerance index for each of the indicators. In all cases the tolerance indices were higher than 0.1 and the FIV lower than 5, as recommended by Kleinbaum, Kupper & Muller (1988), thus refuting the hypothesis of multicolinearity of the formative indicators making up the satisfaction scales for external information sources and attraction of sales promotions. Figure 2 shows the results of the structural model for the two groups of tourists, including the estimated coefficient and t-values resulting from application of the bootstrap resampling procedure to 1000 sub-samples of identical dimensions to the sample size of each of the two groups. Figure 2 - Structural Model: Inexperienced Groups vs. Experienced Group

Por su parte, para los constructos con indicadores formativos sí que es importante valorar la posibilidad de multicolinealidad de los indicadores formativos dado que una elevada colinealidad entre ellos conduciría a problemas en la separación de las diferentes influencias de cada indicador individual sobre la variable latente. Con este fin se calculó en cada grupo el factor de inflación de la varianza (FIV) y el índice de tolerancia para cada uno de los indicadores. En todos los casos los índices de tolerancia fueron superiores a 0,1 y los FIV inferiores a 5 tal y como recomiendan Kleinbaum, Kupper y Muller (1988), de ahí que rechacemos la hipótesis de multicolinealidad de los indicadores formativos que conforman las escalas de

Journal of Marketing Trends - Volume I (February 2011) 57


24366_int_Mise en page 1 22/12/10 12:25 Page58

JOURNAL OF MARKETING TRENDS - TOURISM

satisfacción hacia las fuentes de información externas y atractivo de las promociones de ventas. La figura 2 muestra los resultados del modelo estructural para los dos grupos de turistas, en el que se incluyen los coeficientes y los valores t estimados que resultaron de aplicar el procedimiento de remuestreo Bootstrap a 1000 submuestras de idéntica dimensión al tamaño muestral de cada uno de los 2 grupos. FIGURA 2 - Modelo estructural: grupo sin experiencia vs. grupo con experiencia

Contrario a lo planteado en las hipótesis H1a, H1b, H1c, H1d, H1e, para los más expertos ninguna de las fuentes de información consideradas parece tener una influencia significativa sobre la satisfacción global del turista. Una posible respuesta a este resultado sería que este tipo de turistas ya tienen una experiencia previa con el destino de turismo rural y, por tanto, para ellos no les resulta nada relevante la información que puedan recibir de fuentes externas puesto que tienen un conocimiento suficiente a través de las fuentes internas. Para los más noveles sí que resultan significativas las fuentes de información comerciales impersonales como la publicidad a través de los medios masivos (0,19), las no comerciales impersonales como las noticias al respecto que pueden aparecer en los medios de comunicación (-0,35) y las fuentes de información relacionadas con Internet (-0,20). El hecho de que estas dos últimas relaciones tengan un signo negativo puede deberse a que sus expectativas generadas a través de la consulta de esas fuentes no se hayan visto recompensadas durante la estancia. Teniendo en cuenta que para este grupo las puntuaciones medias de satisfacción eran superiores a 4,6 cabe pensar que consideren que la información recibida por esas fuentes sea insuficiente o inadecuada. A la vista de estos resultados sólo se cumple la hipótesis H1a para las fuentes de información comerciales e impersonales. Asimismo, no se debe rechazar la hipótesis H5 puesto que efectivamente se percibe un efecto moderador de la experiencia previa en las relaciones analizadas.

58 ISSN 1961-7798 - © 2009, International Marketing Trends Conference

On the contrary to the propositions of hypotheses H1a, H1b, H1c, H1d, and H1e, none of the information sources considered seem to have a significant influence on overall satisfaction for the more expert group of tourists. For the more novice tourists, significant influence was exerted by impersonal commercial information sources, such as mass media advertising (0.19), impersonal non-commercial sources such as news items appearing in the media (-0.35), and Internet-related information sources (-0.20). In view of these results, only hypothesis H1a is fulfilled for impersonal commercial information sources. Likewise, hypothesis H5 should not be rejected, since a moderating effect of previous experience on the relations analyzed can be detected. To test hypothesis H2, we analyzed the direct effect of perceived value on satisfaction. This effect was significant and positive for both groups, thus confirming H2 according to previous academic literature on marketing in general and tourism in particular. For tourists with no previous experience, perceived value is an important antecedent of satisfaction (0.72). By comparison, tourist with more experience showed a rather lower direct effect on satisfaction (0.60). Nonetheless, after comparing all the differences between the two groups, none were found to be significant (p>0.05), so that we cannot confirm the moderating influence of previous experience on the relation between perceived value and satisfaction (H6) (comparison of the differences between coefficients was done using the formula proposed by Chin (2000), which takes into account the estimations obtained by the bootstrap resampling technique). Hypothesis H3 considers the positive influence between perceived value of the tourist destination and behavioural intention. To test this hypothesis, we calculated the total effect, given that the perceived value affects intention both directly and also indirectly through satisfaction with the destination. The results confirm the hypothesis both for novice clients (0.56) and for clients that had already visited the destination (0.57). In this case the comparison of differences showed no differences between the two client groups (p>0.05), so that the moderating effect of experience on this relation could not be confirmed (H7). The results confirm the hypothesis H4, given that the paths obtained for both groups were significant (0.46 for novices and 0.27 for experts). Despite some punctual differences, the comparison of differences showed no significant differences between the two groups (p>0.05), so that the influence of previous experience with the destination on the relation between satisfaction and behavioural intention could not be confirmed (H8). Having tested the hypotheses by means of the comparative analysis of structural coefficients, the next and last stage consists in evaluating the predictive capacity of the model. This involves examining the


24366_int_Mise en page 1 22/12/10 12:25 Page59

JOURNAL OF MARKETING TRENDS - TOURISM

Para contrastar la hipótesis H2 se analizó el efecto directo del valor percibido sobre la satisfacción. En ambos grupos éste efecto resultó significativo y positivo, confirmándose H2 en la línea de la literatura académica previa en marketing, en general, y en turismo, en particular. Para los turistas sin experiencia previa el valor percibido es un antecedente importante de la satisfacción (0,72). En comparación a estos, los turistas con más experiencia presentan un efecto directo sobre la satisfacción algo más pequeño en magnitud (0,60). Realizados los contrastes de diferencias entre ambos grupos tales diferencias no resultaron significativas (p>0,05), por lo que no se puede confirmar la influencia moderadora de la experiencia previa en la relación entre el valor percibido y la satisfacción (H6) (el test de diferencias de coeficientes se realizó usando la fórmula propuesta por Chin (2000) que tiene en cuenta las estimaciones obtenidas por la técnica de remuestreo bootstrap). Por otro lado, la hipótesis H3 recoge la influencia directa y positiva entre el valor percibido del destino turístico y la intención de comportamiento. Para testar esta hipótesis se calculó el efecto total dado que el valor percibido afecta a la intención de comportamiento de manera directa y de manera indirecta a través de la satisfacción global hacia el destino. Los resultados confirman esta hipótesis, tanto para los clientes noveles (0,56) como para los clientes que ya han visitado el destino (0,57). En este caso el contraste de diferencias no mostró diferencias entre ambos grupos de clientes (p>0,05), por lo que no podemos confirmar el efecto moderador de la experiencia en esta relación (H7). Los resultados confirman la hipótesis H4 dado que tanto para el grupo de menos expertos como para el de mayor experiencia los paths obtenidos resultaron significativos (0,46 y 0,27, respectivamente). A pesar de las diferencias puntuales observables, el contraste de diferencias no arrojó diferencias significativas entre ambos grupos (p>0,05), por lo que no se puede confirmar la influencia de la experiencia previa con el destino en la relación entre la satisfacción y la intención de comportamiento (H8). Una vez contrastadas las hipótesis mediante el análisis comparativo de los coeficientes estructurales, la siguiente y última fase consiste en la evaluación de la capacidad predictiva del modelo. Para ello se debe examinar el porcentaje de varianza explicada de los constructos dependientes (R2), así como su predictividad a través del indicador Stone-Geisser Q2 que ofrece una medida de la bondad con que los valores observados son reconstruidos por el modelo y sus parámetros (Chin 1998). Se establece que si Q2 > 0 el modelo tiene relevancia predictiva. Analizando los resultados obtenidos para las tres variables dependientes de nuestro modelo (figura 2) se puede afirmar que tanto la satisfacción global hacia el destino

percentage of explained variance of the dependent constructs (R2), as well as their predictivity using the Stone-Geisser Q2 indicator, which provides a measure of the goodness with which the observed values are reconstructed by the model and its parameters (Chin 1998). It is established that if Q2 > 0, the model has predictive relevance. By analyzing the results obtained for the three dependent variables of our model (Figure 2), we can state that both overall satisfaction with the destination (novices: R2=0.60 - Q2=0.40; experts: R2=0.37 - Q2=0.19) and behavioural intention (novices: R2=0.40 - Q2=0.14; experts: R2=0.37 - Q2=0.12) present suitable percentages of explained variance and good predictability in both groups.

DISCUSSION The results show that if the tourist has not visited the destination before, this only has a significant, positive effect on satisfaction with impersonal, commercial (advertising) sources, because in this case the information provided to the individual created expectations about the destination that fit with what he later found to be true. On the basis of these results, we observe the convenience of advertising as an effective instrument of communication in the rural tourism sector. In the case of non-commercial impersonal sources (publicity), the tourist’s satisfaction with them has a negative effect on his overall satisfaction with the destination. The expectations caused by this type of information are higher than the satisfaction obtained on visiting the destination. Although it might be expected that the information provided by a non-commercial source would be true to reality, it may happen that it is perceived as very partial information that does not give a true picture and thus causes excessive expectations in the consumer. Information obtained via Internet creates expectations that, in this case, are not true to reality either. Internet is an information medium with very peculiar characteristics and it allows the individual to compile a large amount of information that in some cases can help him to plan his holiday in detail, but may also cause an information overload (Ahuja & Webster 2001; Eveland & Dunwoody 2001), preventing the consumer from constructing a true image of the holiday destination. For personal commercial information sources (travel agencies, tourist offices and fairs) no significant influence is observed by satisfaction with the source on overall satisfaction. Finally, the influence of satisfaction with noncommercial, impersonal sources (friends and relatives) on overall satisfaction is no significant. This may be due to the fact that, despite this information source being used quite often in this type of tourism, it is nonetheless a very restricted channel that requires a lot of time to extend

Journal of Marketing Trends - Volume I (February 2011) 59


24366_int_Mise en page 1 22/12/10 12:25 Page60

JOURNAL OF MARKETING TRENDS - TOURISM

(noveles: R2=0,60 - Q2=0,40; expertos: R2=0,37 - Q2=0,19) como la intención de comportamiento (noveles: R2=0,40 - Q2=0,14; expertos: R2=0,37 - Q2=0,12) presentan adecuados porcentajes de varianza explicada y una buena predictibilidad en ambos grupos.

DISCUSIÓN Los resultados muestran que si el individuo no ha visitado el destino vacacional con anterioridad únicamente tiene efecto significativo y positivo sobre la satisfacción las fuentes impersonales y comerciales (publicidad), en este caso la información proporcionada al individuo le ha generado unas expectativas sobre el destino que se ajustan a lo que con posterioridad encuentra en el destino vacacional elegido. A partir de estos resultados se observa la conveniencia de la publicidad como instrumento efectivo de comunicación en el sector de turismo rural. En el caso de las fuentes impersonales no comerciales (publicidad no pagada), la satisfacción del turista con esta fuente influye negativamente en la satisfacción global con el destino. Con este tipo de información las expectativas que se genera el individuo son mayores que la satisfacción que le genera el haber visitado el destino. A pesar de ser una fuente imparcial puede ocurrir que sea una información muy parcial y no se ajuste a la realidad provocando expectativas desmesuradas al consumidor. La información que el consumidor obtiene a través de Internet genera unas expectativas que tampoco en este caso se ajustan a la realidad. Internet es un medio de información que tiene unas características muy peculiares y permite al individuo recopilar un gran volumen de información, que en algunos casos, puede ayudarle a planificar detalladamente su estancia vacacional, pero que puede llegar a provocarle sobrecarga de información (Ahuja y Webster 2001; Eveland y Dunwoody 2001), impidiendo que el individuo se construya una imagen real del destino que va a visitar. Para las fuentes de información comerciales y personales (agencias de viaje, oficinas de turismo y ferias), no se observa una influencia significativa de la satisfacción con la fuente con la satisfacción global. Por último, las fuentes no comerciales e impersonales (amigos y familiares) contrario a lo que cabría esperar no es significativa. Este puede deberse a que a pesar de que en este tipo de turismo se utilice bastante esta fuente de información, sin embargo, es un canal de muy poco alcance y que requiere de mucho tiempo para dar a conocer ampliamente los alojamientos y actividades que pueden desarrollarse. Esta investigación pone de manifiesto que cuando se trata de individuos que han visitado en otras ocasiones ese destino vacacional la satisfacción con las fuentes de información no tiene influencia sobre la satisfacción global. Estos resultados proporcionan soporte empírico a la literatura donde se indicó que las fuentes de infor-

60 ISSN 1961-7798 - © 2009, International Marketing Trends Conference

knowledge about the accommodations and activities available. This study shows that for individuals that have visited the holiday destination on other occasions, satisfaction with information sources has no influence on overall satisfaction. These results provide empiric support for the literature describing internal information sources as the main resource for tourists repeating their holiday destination. By analyzing the influence of the tourist’s perceived value of a rural accommodation on his satisfaction with it, as well as the direct influence of satisfaction on loyalty or behavioural intention, we can conclude that there is a positive relation. On examination of whether these relations are conditioned by the fact of it being a first-time visitor or one with previous experience, it was confirmed that the relations are maintained for both type of client, although the strength of such relations is reduced in the case of experienced clients, who will need other incentives to return to that destination, since the market offer such a wide range of tourists destinations from which to choose. Examination of the direct influence of perceived value on behavioural intention has also confirmed that this relation is positive. However, on analyzing where the relation is conditioned by experience with the destination, it was found that it slightly increased in the case of experienced clients. Although the differences between the two groups are not statistically significant, a trend is clearly present, so we may conclude that the perceived value of client with experience of a particular tourist destination is based more on hedonistic benefits related to the desire to explore, self-expression and entertainment, which can increase the intention to return.

MANAGERIAL IMPLICATIONS Concerning the implications for management, the first is the need for managers to know the degree of satisfaction with the different information sources consulted by visitors, as well as the dimensions of perceived value, in order to suitably design their communication policies and promotion of their destinations. Second, it is particularly interesting to mention that managers should not concentrate only on first-time clients, but should continue to attract experienced clients by incorporating new information sources and updating existing ones, in order to improve perception of value and future purchasing behaviour and purchase repetition. In this respect, making use of these results, the managers of rural accommodations should incorporate new incentives and renewed activities, because their aim is not just to attract new clients, but to make former clients return.


24366_int_Mise en page 1 22/12/10 12:25 Page61

JOURNAL OF MARKETING TRENDS - TOURISM

mación internas son el principal recurso para los turistas que repiten su destino vacacional. Al analizar la influencia del valor percibido por el turista de un alojamiento rural en la satisfacción con ese alojamiento turístico, así como la influencia directa de la satisfacción en la fidelidad o la intención de comportamiento, hemos concluido que existe una relación positiva. Al estudiar si estas relaciones se verán condicionadas por el hecho de que se trate de un consumidor que visita el destino por primera vez, o que repita su estancia en el destino, hemos podido confirmar que las relaciones se mantienen para ambos tipos de clientes, aunque la fortaleza de tales relaciones se ve reducida para el caso de los clientes experimentados. Esta reducción en la relación entre el valor percibido y la satisfacción, y también entre ésta y la fidelidad, para el caso de clientes experimentados, en relación a los clientes que visitan el destino por primera vez, nos sugiere un cierto grado de “aburrimiento” con el destino, para los clientes que ya han visitado el alojamiento rural con anterioridad, de forma que su valor percibido sobre el destino tendrá menor influencia en su satisfacción, y lo que es más importante, su intención de comportamiento no estará tan influida por la satisfacción experimentada, sino que necesitará otros incentivos para volver. Cuando estudiamos la influencia directa del valor percibido en la intención de comportamiento, también hemos podido confirmar que esta relación es positiva; sin embargo, al analizar si la relación se verá condicionada por la experiencia con el destino hemos constatado que esta relación aumenta ligeramente para el caso de clientes experimentados. Aunque las diferencias entre los dos grupos no resultan estadísticamente significativas, entendemos que se está poniendo de manifiesto una tendencia, de modo que podemos concluir que el valor percibido de un cliente experimentado con un destino turístico se basa más en beneficios hedonistas relacionados con el deseo de explorar, la autoexpresión y el entretenimiento, que pueden hacer que la intención de volver se vea incrementada.

LIMITATIONS AND FURTHER RESEARCH An important limitation of this study concerns the impossibility of differentiating between different types of accommodation to be found in the rural sector, which has prevented us from making a differentiated, specific profile for each type of accommodation. Finally, another important limitation is the lack of consideration of variables such as perceived quality, which have been analyzed as antecedents of satisfaction and prediction of the client’s future behaviour. Acknowledgments: This study was carried out thanks to funding received for the Aula de Emprendedores research project – Andalusian Agency for Innovation and Development - University of Granada, code Z162.

IMPLICACIONES PARA LA GESTIÓN Respecto a las implicaciones para la gestión que surgen de esta investigación cabe citar las siguientes. La primera es la necesidad de que los gestores conozcan el grado de satisfacción de las diferentes fuentes de información consultadas por los turistas, así como las dimensiones que conforman su valor percibido al objeto de diseñar cuidadosamente las políticas de comunicación y promoción de sus destinos turísticos. En segundo lugar, es importante que los gestores turísticos no centren toda su atención sobre los nuevos clientes y traten de atraer a clientes con experiencia en turismo rural incorporando nuevas fuentes de información y actualizando las existentes, al objeto de mejorar la

Journal of Marketing Trends - Volume I (February 2011) 61


24366_int_Mise en page 1 22/12/10 12:25 Page62

JOURNAL OF MARKETING TRENDS - TOURISM

percepción del valor, el comportamiento de futuro de visita y la repetición de dicho comportamiento. En relación con lo anterior, entendemos que es necesario que los gerentes de los alojamientos rurales incorporen nuevos incentivos, y actividades renovadas, puesto que su objetivo será, no solo atraer a nuevos clientes, sino conseguir que quienes ya han visitado el destino vuelvan a hacerlo.

LIMITACIONES Y FUTURA INVESTIGACIÓN Una primera limitación de este trabajo tiene que ver con la imposibilidad de diferenciar entre diferentes tipos de alojamientos en turismo rural, lo que nos hubiera permitido detectar diferentes tipos de perfiles por tipo de alojamiento. Otra importante limitación tiene que ver con la no consideración en la investigación de variables tales como la calidad percibida, que ha sido analizada en investigaciones previas como un importante antecedente de la satisfacción y del comportamiento futuro. Agradecimientos: Este estudio ha sido llevado a cabo gracias a la financiación recibida del Proyecto de Investigación “Aula de Emprendedores” concedido por la Agencia Andaluza de Innovación y Desarrollo, código Z162.

Agradecimientos: Este estudio ha sido llevado a cabo gracias a la financiación recibida del Proyecto de Investigación “Aula de Emprendedores” concedido por la Agencia Andaluza de Innovación y Desarrollo, código Z162.

Beatty, S. and Smith, S. (1987), “External search effort: An investigation across several product categories”, Journal of Consumer Research, 14, 83-95. Bettman, J.R. and Sujan, M. (1987), “Effects of framing on evaluation of comparable and non-comparable alternatives by expert and novice consumers”, Journal of Consumer Research, 14 (September), 141-154. Bigné, J.E., Sánchez, M.I. and J. Sánchez, (2001), “Tourism image, evaluation variables and after purchase behaviour: inter-relationship”, Tourism Management, 22, 607-616. Capella, L. and Greco, A. (1987), “Information sources of elderly for vacation decisions”, Annals of Tourism Research, 14, 148-151. Castañeda, J.A., Frías, D.M. and Rodríguez, M.A. (2007), “The influece of the Internet on destination satisfaction”, Internet Research: Electronic Networking Applications and Policy, 17(4), 404-420. Chen, J. and Gursoy, D. (2001), “An investigation of tourists destination loyalty and preferences”, International Journal of Contemporary Hospitality Management, 13(2), 79 Chin, W.W. (1998), “The partial least squares approach to structural equation modelling”, In Modern methods for business research G.A. Marcoulides, ed., 295-336. Mahwah, NJ: Lawrence Erlbaum Associates, Publisher. Chin, W.W. (2000). Frequently Asked Questions – Partial Least Squares & PLS-Graph. Home Page [On-line]. <http://disc-nt.cba.uh.edu/chin/plsfaq.htm>.

Referencias / Bibliography

Cronin, J.J. and Taylor, S.A. (1992), “Measuring service quality: a reexamination and extension”, Journal of Marketing, 56, 55-68.

Ahuja, J. and Webster, J. (2001), “Perceived disorientation: an examination of a new measure to assess web design effectiveness”, Interacting with Computers, 14, 15-29.

Cronin, J.J.Jr; Brady, M.K. and Hult, G.T.M. (2000), “Assessing the effect of quality, value and customer satisfaction on consumer behavioral intentions in services environment”, Journal of Retailing, 76, 2, 193-218.

Alba, J.W., and Hutchinson, J.W. (1987), “Dimensions of consumer expertise”, Journal of Consumer Research, 13(4), 411-454.

DeLone, W.H. and McLean, E.R. (1992), “Information systems success: The quest for the dependent variable”, Information Systems Research, 3(1), 60-95.

Baker, D.A. and Crompton, J.L. (2000), “Quality, satisfaction and behavioural intentions”. Annals of Tourism Research, 27(3), 785-780.

Eveland, W.P. and Dunwoody, S. (2001), “User Control and Structural Isomorphism or Disorientation and Cognitive Load”, Communication Research, 28(1), 48-78

Baloglu, S. (2000), “A path-analytical model of visitation intention involving information sources, socio-psychology motivations and destinations images”, In Consumer Psychology of Toursim Hospitality and Leisure, A. Woodside, G. Crouch, J. Mazanec, M. Oppermann y M. Sakai, eds., 63-90. New York: CABI.

62 ISSN 1961-7798 - © 2009, International Marketing Trends Conference

Fornell, C. (1992), “A national customer satisfaction barometer: The swedish experience”, Journal of Marketing, 56 (2), 6-21.


24366_int_Mise en page 1 22/12/10 12:25 Page63

JOURNAL OF MARKETING TRENDS - TOURISM

Fornell, C. and Larcker, D.F. (1981), “Evaluating structural equation models with unobservable variables and measurement error”, Journal of Marketing Research, 18, 39-50.

Peterson, R.A., Balasubramanian, S. and Bronenberg, B.J. (1997), “Exploring the implications of the Internet for consumer marketing”, Journal of the Academy of Marketing Science, 25 (4), 329-346

González Gallarza, M. y Gil Saura, I. (2006), “Value dimensions, perceived value, satisfaction and loyalty: an investigation of university students’ travel behaviour”, Tourism Management, 27(3), pp. 437452.

Petrick, J.F. (2004), “First timers’ and repeaters’ perceived value”, Journal of Travel Research, 43(1), 29-38.

González Gallarza, M.; Arteaga Moreno, F. and Gil Saura, I. (2005), “La investigación del comportamiento del turista: Propuesta de un modelo explicativo de la evaluación postcompra”, XVII Encuentro de Profesores de Marketing, Madrid. Groen, G.J. and Patel, V.L. (1988), “The relationship between comprehension and reasoning in medical expertise”, In The nature of expertise M.T.H. Chi, R. Glaser, y M.J. Farr, eds., 287-310, Hillsdale: Lawrence Erblaum Associates Hall, D. (2004), “Rural tourism Development in southeastern Europe: transition and search for sustainability”, The International Journal of Tourism Research, 6(3), 165-176. IET, Instituto de Estudios Turísticos (2007), Turismo Interior en Andalucía. Madrid: Ministerio de Industria, Turismo y Comercio. Johnson, E.J. (1988), “Expertise and decision under uncertainty: performance and process”, In The nature of expertise M.T.H. Chi, R. Glaser y M.J. Farr, eds., 209-228, Hillsdale: Lawrence Erblaum Associates. Kleinbaum, D.G., Kupper, L.L. and Muller, K.E. (1988), Applied Regression Analysis and Other Multivariables Methods. Boston: PWS-KENT Publishing Company. Lee, Ch., Lee, Y. and Lee, B. (2005), “Korea s destination image formed by the 2002 world cup”, Annals of Tourism Research, 32 (4), 839-858. McKinney, V., Yoon, K. and Zahedi, F. (2002), “The measurement of web-customer satisfaction: An expectation and disconfirmation approach”, Information System Research, 13(3), 296-315. Molina, A. and Esteban, A. (2006), “Tourism brochures. Usefulness and Image”, Annals of Tourism Research, 33(4), 1036-1056.

Petrick, J.F. and Backman, S.J. (2002), “An examination of the construct of perceived value for the prediction of golf travelers’ intentions to revisit”, Journal of Travel Research, 41(1), 38-45. Rao, A.R. and Sieben, W.A. (1992), “The effect of prior knowledge on price acceptability and the type of information examined”, Journal of Consumer Research, 19, (September), 256-270. Schul, P. and Crompton, J.L. (1983), “Search behaviour of international vacationers: Travel-specific lifestyle ans sociodemographic variables”, Journal of Travel Research, 21, 25-31. Sirakaya, E. and Woodside, A. (2005), “Building and testing theories of decisión making by travelers”, Tourism Management, 26, 815-832. Spence, M.T. and Brucks, M. (1997), “The moderating effects of problem characteristics on experts’ and novices’ judgments”, Journal of Marketing Research, 34(may), 233-247. Spreng, R.A., MacKenzie, S.B. and Olshavsky, R.W. (1996), “A reexamination of the determinants of consumer satisfaction”, Journal of Marketing, 60(july), 15-32. Szymanski, D.M. and Hise, R.T. (2000), “E-satisfaction: An initial examination”, Journal of Retailing, 76(3), 309-322. Tian-Cole, S., Crompton, J.L. and Willson, V.L. (2002), “An empirical investigation of the relationships between service quality, satisfaction and behavioural intentions among visitors to a wildlife refuge”, Journal of Leisure Research, 34(1), 1-24. Um, S., and Crompton, J. (1990), “Attitude determinants in tourism destination choice”, Annals of Tourism Research, 17, 432-448. Wakefield, K.L. and Barnes, J.H. (1996), “Retailing hedonic consumption: A model of sales promotion of a leisure service”, Journal of Retailing, 72(4), 409-427.

Oh, H. (1999), “Service quality, customer satisfaction, and customer value: A holistic perspective”, International Journal of Hospitality Management, 18(1), 67-82. Oliver, R. and Swan, J. (1989), “Consumer perceptions of interpersonal equity and satisfaction in transactions: A field survey approach”, Journal of Marketing, 53(2), 21-35.

Journal of Marketing Trends - Volume I (February 2011) 63


24366_int_Mise en page 1 22/12/10 12:25 Page64

64 ISSN 1961-7798 - Š 2009, International Marketing Trends Conference


24366_int_Mise en page 1 22/12/10 12:25 Page65

JOURNAL OF MARKETING TRENDS - ADVERTISING

The Effectiveness of Slow and Fast paced TV Commercials: some Experimental Empirical Results

The main objective of this paper is analyse the influence of age and experience of fast- and slow visual information processing on the brand awareness and comprehension of TV commercials. The research results indicate that age has a negative influence on the awareness of the brand shown in TV commercials. The research results also indicate that consumers with significantly more experience with processing of fast visual information have a significantly higher score on the comprehension of ‘fast’ TV commercials than consumers with less experience in processing fast visual information. This suggests that if the advertiser has to choose between a fast and a slow paced TV commercial, the best choice will normally be a fast paced commercial. Keywords: Information Processing, Paced Advertising, TV commercials, Comprehension

> Irene Roozen HUBrussels - Associated with Catholic University Leuven, Belgium Irene.Roozen@hubrussel.be

> Kristin Blondé Kristin.blonde@vmmtv.be

Journal of Marketing Trends - Volume I (February 2011) 65


24366_int_Mise en page 1 22/12/10 12:25 Page66

JOURNAL OF MARKETING TRENDS - ADVERTISING

Introduction and Overview of the Literature One of the consequences of our changing daily lives is that the amount of (visual) information we are exposed to is increasing. Statistics show that the number of minutes we are watching television, playing computer games and surfing on the internet and using our mobile phone is still growing (Eurostat, 2007; Federale Overheidsdienst Economie, 2006; Federaal Planbureau, 2006; Vandewater et al., 2007). Consequently, our exposure to commercial information is also increasing (Studiedienst van de Vlaamse Regering, 2006). It is, therefore, important to analyse if we can still process and comprehend all these commercial communications and if there are differences between viewers.

Overview of the literature The overview of the literature will focus on (1) the comprehension of fast and slow paced TV commercials and (2) the influence of age and experience with visual information processing on the comprehension of fast and slow TV commercials.

Comprehension of fast and slow paced TV commercials Comprehension of commercial communications is often defined as understanding what goes on in people’s minds when they are exposed to an advert or any form of marketing communication in response to it. This can include all kind of reactions (Percy and Rossiter, 2001). For measuring comprehension, research often refers to the ACCA or DAGMAR model of Colley (1961). In the psychology literature the processing of the information is widely analysed and the results are subsequently used by marketers. The literature shows that the hierarchy-ofeffects models have dominated the marketing communications for the last decades. According to these models, consumers go through three different stages in responding to marketing communications. The first stage is the cognitive stage, followed by the affective stage and finally the conative stage (Barry and Howard, 1990). This research is concentrated on the effect of marketing communications on the cognitive stage. During this stage consumers engage in mental (thinking) processes which lead to brand awareness and comprehension of commercial communicated. According to the limited capacity model of Lang (2000), comprehension of a television message involves the continuous and simultaneous operation of three parallel cognitive sub processes: encoding, storage and retrieval. This makes television viewing a complex and difficult cognitive task (Lang et al., 1999). New information from television messages is continuously attended

66 ISSN 1961-7798 - © 2009, International Marketing Trends Conference

to, encoded into working memory, processed, and stored. Previously held information, required to understand the message, is concurrently retrieved, associated with new information, and stored again. Information encoded earlier of the message is being stored and later information is being encoded. A viewer of a commercial message cannot encode and store all the information from the message and will therefore select which information in the message to encode, to process and to store. Consequently, the viewer, the medium and the content all affect how the message is processed. The processing of visual commercial information is becoming more and more important because of the increasing speed and complexity of visual information nowadays. The modern advertising industry uses more fast paced and complex messages (Rossiter and Bellman, 2005). Commercials on television are more and more produced in a ‘Music Television-style’: Commercials are presented in a rapid paced, multi-visual mode and often accompanied with upbeat music and other special effects. These commercials are designed in the hope of finding a perfect mix of visuals and advertisement copy that will “cut through the clutter and gain the attention of the television viewing audience” (Tse and Lee, 2001). Advertisers intuitively believe that the ‘faster’ the better, which means that a higher speed at which the information is visually presented to the viewer for cognitive processing – the commercial pace – is better. However, there is very little academic research on this subject (Bolls et al. 2003). Rossiter and Bellman (2005) describe that slow- and fast paced TV commercials are equally good at holding attention, however, fast-cut commercials produce lower recall, and therefore less learning of the central message content. Bolls et al. (2003) found that the skin conductance responses are elicited significantly more in the fast paced advertisements than in slow-paced advertisements. The memory/recall score in their research was also significantly higher for the fast paced advertisements compared to the slow paced advertisements. In the Web environment, fast-animation ads also generate significantly greater arousal than slow-animation ads. However no significant differences were found for ad recognition scores (Sundar and Kalyanaraman, 2004). According to the limited capacity theory of processing, the viewers’ cognitive capacity for processing advertising stimuli is subject to limitations (Lang et al. 1999; Lang, 2000). Due to viewers’ limited processing capacity, one may expect that the fast paced commercials in some cases cause information overload for viewers and that they may not always be able to comprehend the numerous bits of visual information that are presented rapidly. On the other hand, the stimuli in fast paced commercials are stronger – the level of arousal is


24366_int_Mise en page 1 22/12/10 12:25 Page67

JOURNAL OF MARKETING TRENDS - ADVERTISING

significantly higher – which can result in a higher attention level which, in turn, can ameliorate the comprehension level. Lang et al. (1999) have found that pacing has generally increased recognition for calm messages, but resulted in a decrease for arousal messages, because the combination of arousal content and fast pace pushes the system into cognitive overload. Scott and Batra (2003) consider the discrepancy between the ‘real’ psychical world and the mental representations in the domain of visual processing and its implications for visual advertising. One of their main conclusions is that - contrary to Bolls et al. (2003) - broadcasting of ‘fast’ visual information has a negative effect on the awareness of brands shown in advertisements.

The influence of age and experience with visual information processing on the comprehension of fast and slow TV commercials Science suggests that age and experience can influence information processing. According to medical science, today's students think and process information fundamentally differently than their predecessors (Prensky, 2001). Psychology research has also indicated that age has a significant influence on the information processing of commercial information (Scharps, 1998; Scharps and Gollin, 1998; Salthouse 1993; Salthouse, 1991). Older consumers have a lower score on brand awareness and comprehension of commercial communications. Salthouse (1991) has used the term ‘reduced processing resources’. However, he has found that the influence of age on cognition is not direct but indirect, whereby ‘speed of information processing’ was found as an explanation of the reciprocal correlation. Conversely, ‘knowledge and experience with the subject’ has a positive influence on information processing. This can also influence the relationship between age and cognition (Mackay and Abrams, 1996; Kolb and Wishaw, 2001). They suggest that experience with fast visual exposures can reduce this negative effect. This means that after seeing fast visual information, consumers with significantly more experience with fast visual information have a significantly higher score on brand awareness and comprehension of ‘fast’ TV commercials than consumers with less experience with fast visual information (and vice versa). The literature suggests that age tends to be negatively correlated with cognition, and experience with information positively. The literature of fast and slow paced commercials suggests that fast ‘visual’ information experiences may have conflicting influences on the brand awareness and comprehension and that as a consequence, the net effect cannot be determined a priori.

Research Questions Based on the introduction and the literature review, this research takes the speed of pacing of the TV commercials, the experience of visual information processing of the viewer and the age of the viewer into account. The main research questions are concerned with the influence of age and experience of visual information processing on brand awareness and comprehension of TV commercials as results of information processing. Q1: Does age influence the brand awareness of TV commercials? Q2: Is comprehension of fast paced TV commercials stronger than for slow paced TV commercials? Q3: Does experience with watching ‘fast’ paced visual information processing influence the comprehension of ‘fast’ paced TV commercials (and vice versa)?

Data Collection Two different experimental groups were exposed to the same TV- soap series with the same TV commercials. The first group consists of 116 Dutch speaking students of Flanders (Belgium) between 15-18 years: the digital ‘natives’ (M=16.91, SD=0.82; 44% male). The second group consists of 80 Dutch speaking subjects between 45-54 years: the digital ‘immigrants’ (M=49.58, SD=3.35; 50% male) (see Prensky, 2001). Prensky (2001) has introduced the terms digital natives and digital immigrants, because differences are related to whether a person grew up with or without fairly accessible electronic media. Those who did have been referred to by Prensky (2001) as digital natives while the rest who function within an ITC world are digital immigrants. As a result many major companies have adopted the vernacular of digital natives and digital immigrants (see for example the way Apple pushes its technologies into the education sector in the U.S.). For both age samples, the subjects were recruited at random and asked to watch the program (the TV-drama series with the TV commercials) on a monitor and fill out a questionnaire about the programme. Before the programme was presented, the subjects were told that the research investigated the effectiveness of product placements. This was done to avoid a too strong focus on the TV commercials on the part of the respondents. A broadcast of an existing TV-drama series was used to maximize the external validity. The programme is not specifically targeted to a specific age group but widely appreciated in Flanders (Dutch speaking part of Belgium) by all age groups (CIM, 2009). Eventual carry-over effects of mood induced by the programme are taken into account by using the same programme for both experimental groups. The 6 TV commercials which were shown had - at time of the experiment - never been broadcasted on the Belgian TV-channels.

Journal of Marketing Trends - Volume I (February 2011) 67


24366_int_Mise en page 1 22/12/10 12:25 Page68

JOURNAL OF MARKETING TRENDS - ADVERTISING

After 10 minutes the TV programme was interrupted by 3 different TV commercials, then the TV programme continued for another 10 minutes and at the end another 3 different TV commercials were shown. The products shown in the TV commercials belong to different product categories. The TV commercials were a mixture of 3 ‘fast’ and 3 ‘slow’ paced TV commercials, whereby the sequence of the fast and slow paced TV commercials was mixed. The classification of the TV commercials into ‘fast’ and ‘slow’ was based on the ‘speed’ and the ‘complexity’ of the TV commercials. A ‘fast’ paced TV commercial was defined as a TV commercial with a relatively high quantity of images in sequences, a relatively high score on discontinuity of sequences and a relatively high number of sequences. The classification was judged individually by 4 communication marketing experts. Their classification was similar to the theoretical classification of the 6 TV commercials on the basis of the amount of cuts of every commercial (the video scene changes that occur on average every half-second or faster). Rossiter and Bellman (2005) define that the average number of cuts in a 30-second TV commercial is 13, more than this is defined as a ‘fast cuts’ TV commercial. Immediately after exposure to the total transmission (about 23 minutes), the subject was asked to turn the page and to fill out the written questionnaire.

Measurement Instruments The questionnaire started with questions about the subjects’ knowledge of the different brands which figured in the TV programme in order to keep up the invented objective of the research (product placements) and brand recall of the different TV commercials. On the second page questions were asked about the recognition of the different brands in the TV commercials. On the third page several questions about the individual TV commercials were asked. For each TV commercial the respondent was asked if he had seen this commercial before (and how often), if the respondent had the impression that the TV commercial had a high variety of different images/pictures (Likert 7-point scale), if the respondent found the ‘speed’ of the TV commercial high (if the alternation of images/pictures was high, on a Likert 7-point scale) and if the respondent had difficulties to understand the TV commercial because of the language which was spoken in the commercial (for 5 out of 6 TV commercials the spoken language was English instead of Dutch). The comprehension of the TV commercial was measured by two aspects. First, the understanding of the message of the TV commercial was measured, 4 different messages were given while only 2 messages were correct. Second, a question was asked concerning the story of the TV commercial: what do you think is the story of the producer for this TV commercial? Also here, the

68 ISSN 1961-7798 - © 2009, International Marketing Trends Conference

responses of only 2 out of 4 stories were correct. At the end of the questionnaire, subjects were asked the amount of time they were watching television and the TVchannels they were watching (with their frequency). They were also asked the amount of time they spent on the internet (and reasons for using internet) and the time they spent on playing computer games during the week.

Research Results Before analysing, the data were screened for respondents who had already experienced the TV commercials and for respondents who had problems in understanding the language which was spoken during the TV commercials. Having already experienced the TV commercial can positively bias the results for comprehension. In total 1176 commercial viewer combinations could be analysed. Before investigating the research questions, the research design for the slow and fast paced TV commercials which we were included in this study were tested. For this purpose, it was analysed if the TV commercials used in the experiment were also perceived as ‘fast’ and ‘slow’ paced TV commercials by the different experimental groups. In Table 1 the results for the different TV commercials which were selected for the experiment were tested for their perceived slow/fast value. In this table the total score for the 3 ‘fast’ and the 3 ‘slow’ paced TV commercials is given (which were given on a 7-point Likert scale individually)

Table 1 - Perception of “speed” of the slow and fast paced TV commercials

The research results in Table 1 indicate that both experimental groups perceive the 'slow' paced TV commercials as commercials with a lower amount of different images and a lower speed value of the broadcasting than the fast paced TV commercials. However, the digital immigrants perceive the 'slow’ TV commercials as significantly 'faster' than the digital natives, while for the ‘fast’ TV commercials this result is not found. The perception of the number of images and of the speed of the different images differs significantly between the two


24366_int_Mise en page 1 22/12/10 12:25 Page69

JOURNAL OF MARKETING TRENDS - ADVERTISING

groups (t-value 5.59, p<0.001; t-value 5.83, p<0.001). For the ‘fast’ TV commercials this result is not found. Only the perception of the speed of the different images is significantly different (t-value 3.432, p=0.001). The perception of the number of images in both groups is relatively high and does not differ significantly between the groups. Note that the digital natives differentiate much more strongly between slow and fast paced TV commercials than the digital immigrants. A t-test comparing the scores for ‘slow’ and ‘fast’ indicates that digital immigrants barely distinguish between the categories. The mean difference for digital natives (between 16.42 and 9.66) for the perceived value for the amount of images is significant (value 6.78, p<0.001) and for the perceived value for speed of different images is 4.98 (difference between 14.82 and 9.83) (p<0.001). The mean difference for digital immigrants for the perceived value for amount of images is 2.79 (p<0.001) and 2.18 (p<0.001) for the perceived value for speed of different images. This may be related to the greater experience of digital natives with visual information processing, which allows them to distinguish different types more clearly. This is borne out by the fact that the experience of visual information - ‘watching TV, internet and playing computer games’ - on a daily basis is significantly different between the two experimental groups (see Table 2). The composition of the visual information consumed is also different. For example, digital immigrants indicated that they use internet for reading journals and obtaining textual information whereas the digital natives use it significantly more often for msn, watching video clips and computer gaming (chi-square 115.16; d.f. 5, p-value <0.001). On average the research results on consumption of visual information processing show that the digital natives spend almost 5 hours a day watching TV, internet and playing computer games while the digital immigrant spends only 2.5 hours a day. It should, however, be pointed out that in reality these groups might carry out some of these activities in parallel e.g. playing a computer game, watching TV and surfing on the internet, and using their mobile phone. This would mean that the actual ‘total’ time spent watching on ‘visual information’ could be lower. Table 2 - Average experience of visual information by digital natives and digital immigrants (s.d)

These results are consistent with the literature which suggests that differences are related to whether a person grew up with or without fairly accessible electronic media (Prensky, 2001). This is further illustrated by the fact that the differences are much larger for ‘new’ technologies than for old technology (e.g. TV). Table 3 presents the research results for Q1. Conform the expectations, the brand recall and brand recognition is significantly higher for the digital natives than for the immigrants. Table 3 - Average score on Brand Awareness for the 6 TV commercials (s.d.)

The above analysis is based on an ‘exogenous’ categorisation of the sample population into two groups on the basis of age (here, digital natives, 15-18 years, digital immigrants 45-54 years). This assumes that digital natives have more experience and this is confirmed – on average – by the results presented in Table 2 and higher awareness scores (see Table 3). To corroborate this finding, an endogenised categorisation on the basis of a cluster analysis is analysed. For this purpose, the respondents’ answers with regard to the nine different broadcasting stations (fast e.g. music and commercial channels, and slow paced stations e.g. national and non-commercial channels) which they indicated they watched (always, very often, rarely, never) were analysed. On the basis of a K-means cluster analysis the respondents can be categorised in two groups with different experiences of visual information processing. Group I (N=113 respondents) consists of respondents which are most of the time watching the ‘slow’ broadcasting stations (often national and the not-commercial channels, average age 36 years). Group II (N=83) consists of respondents exposed to the ‘fast’ broadcasting stations (e.g. music channels, channels, average age 23 years). Both groups differ significantly in respect of their use of visual information channels, but not with regard to time spent watching TV (group I: 23.06 hours a week; group II: 24.42 hours a week; t-value 1.09, p-value 0.277), surfing on the internet (Group I: 16.53 hours a week; Group II: 19.24 hours a week; t-value 1,71, p-value 0.089) or playing computer games (Group I: 10.11 hours a week; Group II: 9.94 hours a week; t-value 0.175, p-value 0.861). The two groups did not differ significantly for gender (chi-square 1.73; p-value 0.189). Note that while there is a difference in age, these groups are different from the digital natives and digital immigrants

Journal of Marketing Trends - Volume I (February 2011) 69


24366_int_Mise en page 1 22/12/10 12:25 Page70

JOURNAL OF MARKETING TRENDS - ADVERTISING

defined above (compare time spent on visual information processing with the results in Table 2). Group I consists of 43.4% of digital natives while group II exists of 80.7% of digital natives. In Table 4 the research results for Q2 and Q3 are given. In the questionnaire, comprehension of the TV commercial was distinguished into (1) comprehension of the commercial message of the TV commercial and (2) comprehension of the story of the TV commercial. The research results of both scores of comprehension and the composite score (total score for ‘1’ and ‘2’) of comprehension of the TV commercial are given in percentage of comprehension for the different TV commercials and for the different groups of experiences with visual information processing. For example, the messages of the slow paced TV commercials shown to the total group were well understood in 40.69% of the viewer commercial combinations. However, group I (the group with significant more experience with slow visual information) only understood 37.52% of the TV commercials well and group II understood it significant better with 45.29% (p=0.019).

Table 4 - Comparison of comprehension scores for slow and fast paced TV commercials for groups I with slow - and group II with fast visual information processing experience

For Q2, the research results of Table 4 indicate that the total comprehension score of fast paced TV commercials (41.42%) is stronger than for slow paced TV commercials (38.37%). However, this result is not significantly stronger at a reliability level of 95% (t-value 1.806; p=0.071). This is due to the fact that there is no significant difference between the comprehension score of the message of the slow and the fast paced TV commercials (t-value 0.717, p=0.474). Note that the comprehension of the story of the TV commercial is significantly better for the fast paced TV commercials (43.87%) than for the slow faced TV commercials (36.05%) (t-value 3.545, p<0.001).

70 ISSN 1961-7798 - © 2009, International Marketing Trends Conference

This can be explained by the stimuli in fast paced commercials which are stronger – the level of arousal is significantly higher – which result in a higher attention level which, in turn, can ameliorate the comprehension level of the story of the message. Also the research of Bolls et al. (2003) has shown that fast-paced advertisements result in significantly greater viewer arousal than slow-paced advertisements. The research results for Q3 are in the last three columns of Table 4. For slow paced TV commercials the research results in Table 4 indicate that the total comprehension score is higher for the group with fast visual information processing experience (39.71% compared to 37.45%). However the difference for the total comprehension score for the different experience groups is not significant for the slow paced TV commercials (t-value 0.919, p-value 0.358). The comprehension of the message of the slow paced TV commercials is significantly better for the group with more experience with fast visual information processing experience (t-value 2.355, p=0.019) whereas the story is better understood (but not significantly) by the group with significantly more experience with slow visual information processing (1.049, p=0.295). Notable, the total comprehension score of fast paced TV commercials (message plus story) is significantly different for both groups. The total comprehension score is significantly higher for the group with significantly more fast visual information processing (46.47%) than for group with slow visual information processing (37.82%). Here, the message and the story of the fast paced TV commercials is significantly better understand by viewers with significantly more experience with fast visual information processing. This suggests that experience with watching ‘fast’ paced visual information significantly influence the comprehension of ‘fast’ paced TV commercials. Note that there is a much bigger difference between the comprehension of fast and slow paced TV commercials for group II with fast visual information experience (total comprehension score for slow is 39.71% and for fast is 46.47%, t-value 2.592, p=0.010) than for group I with slow visual information experience (total comprehension score for slow is 37.45% and for fast is 37.82%, t-value 0.167, p=0.868). Table 4 shows that the comprehension score of the message is higher for the slow paced TV commercials for both groups. The comprehension score of the message for slow paced TV commercials is 37.52% for group I; for fast paced this drops to 35.15% (t-value 0.772, p=0.440). The comprehension score of the message of the fast paced commercials for group II is 44.31% but the comprehension score of the message for group II for slow paced TV commercials is 45.29% (t-value 0.259, p=0.796). However, for both groups, the comprehension score of the story is higher for the fast


24366_int_Mise en page 1 22/12/10 12:25 Page71

JOURNAL OF MARKETING TRENDS - ADVERTISING

paced TV commercials. This suggests that if the message of the commercial is very important e.g. a new feature of an existing product, a characteristic of the product, a slow TV commercial will be more effective. Conversely, if the story of the TV commercial is very important, e.g. for image building of the product, one should produce a fast paced TV commercial. If both message and story are equally important, marketers should select a fast paced TV commercial because this will have a bigger overall effect, even irrespective of the visual information processing experiences of the subjects.

Conclusions and limitations On the basis of our research results we can conclude that if an advertiser has to choose between a fast and a slow paced TV commercial, the best choice will normally be a fast paced commercial. Only if the message of the TV commercial is of major importance e.g. new feature or characteristic of a product, slow paced commercials should be used. The research results also indicate that age has a significant negative influence on the brand awareness of the fast and slow paced commercials and that experience of fast- and slow visual information processing influences the comprehension of the commercials. In particular, experience of fast visual information processing leads to a significantly better comprehension of fast paced TV commercials. For age, the research results show that the digital immigrants (average age 50 years) barely distinguish between slow and fast paced TV commercials whereas the digital natives (average age 17 years), probably because of their experiences, clearly differentiate between the two types. The results also show that the consumpion of visual information of the digital natives differs significantly from the digital immigrants. The hours the natives watch TV, play games and use the Internet is significantly higher compared to the digital immigrants. Whereby, the natives use the Internet for msn, watching video clips and computer gaming and the immigrants use it for reading journals and obtaining textual information. The awareness scores (recall and recognition) is also significant higher for the digital natives. As an experiment, this study was subject to several limitations. First, the TV commercials were viewed in conditions isolated from a surrounding media content. Second, the subjects were asked to watch the program on a monitor which was not a natural environment for television viewers. Third, individual differences e.g. experiences with the products, and product category of the brands which were shown in the TV commercials and the cultural background of the subjects were not taken into account in this study. However, two different ages groups were investigated for their experiences with visual information processing and the comprehension level of the TV commercials. These limitations notwithstanding,

the current study sheds additional light on the influence of age and experience of fast- and slow visual information processing on the brand awareness and comprehension of TV commercials, whereby the ‘faster’ is better’ motto of the advertisers can not be denied.

Bibliography Bolls, Paul, D., Darrel D. Muehling and Kak Yoon (2003), The effects of television commercial pacing on viewers’ attention and memory, Journal of Marketing Communications, 9, 17-28. CIM (2009), Centrum voor informatie over de media, Hoogste kijkcijfers van het jaar, http://www.cim.be/tele/nl/d/db.html De Pelsmacker, P., M. Geuens, J. Van den Bergh (2004), Marketing Communications, A European Perspective, second edition, Prentice Hall. Barry, Thomas F. and Howard, Daniel J.(1990) Review and Critique of the Hierarchy of Effects in Advertising, International Journal of Advertising; Vol. 9 Issue 2, p121-135, 15p Colley, R. H. (1961) Defining Advertising Goals for Measured Advertising Results. New York, NY: Association of National Advertisers. Eurostat (2007) Europe in Figures, Eurostat Yearbook 2006-2007 with CD-rom, Eurostat: European Commission, Belgium. Federale Overheidsdienst Economie, http://statbel.fgov.be/figures/d34_nl.asp, 2006, Online.

2006, January

Studiedienst van de Vlaamse Regering, (2006), http://aps.vlaanderen.be/statistiek/ publicaties/pdf/VRIND/vrind2000/media.pdf, January 2006 Federaal Planbureau (2006) http://www.plan.be/websites/ferado/nl/html_books /ferado/ 2nl10.html, January 2006 Kolb,B., Whishaw, I.(2001), An Introduction to Brain and Behavior, second edition, New York, Worth Publishers. Lang, Annie, Paul Bolls, Robert F. Potter and Karlynn Kawahara (1999), The effects of production pacing and arousal content on the information processing of television messages, Journal of Broadcasting & Electronic Media, 43 (4), 451-475.

Journal of Marketing Trends - Volume I (February 2011) 71


24366_int_Mise en page 1 22/12/10 12:25 Page72

JOURNAL OF MARKETING TRENDS - ADVERTISING

Lang, A., (2000), The limited capacity model of mediated message processing, Journal of Communication, 50, 1, 46-70. Lavidge, Robert J. and Steiner, Gary A. (1961) A Model for Predictive Measurements of Advertising Effectiveness, Journal of Marketing, Oct61, Vol. 25 Issue 6, p59, 4p Mackay, D.G. and L., Abrams (1996), Language memory and aging: Distributed deficits and the structure of new-versus-old connections, Handbook of the Psychology of Aging, fourth edition, San Diego, Academic press. Percy, L; J.R. and Rossiter R. E (2001), Strategic Advertising Management, Oxford, Oxford University Press. Petty, R. E. and Cacioppo, J. T. (1986a). Communication and Persuasion: Central and peripheral routes to attitude change. New York: Springer. Petty, R. E. and Cacioppo, J. T. (1986b). The Elaboration Likelihood Model of Persuasion. In L. Berkowitz (Ed.), Advances in experimental social psychology, New York: Academic. Prensky, M. (2001), Digital natives, Digital immigrants, On the Horizon, NCB University Press, 9, 5, October. Prensky, M. (2001), Digital natives, Digital immigrants, Do they really think differently?, On the Horizon, NCB University Press, 9, 6, December. Rossiter, J.R., S., Bellman (2005), Marketing Communications: theory and applications, Australia, Pearson Education Australia. Salthouse, T.A. (1991), Theoretical perspectives on cognitive aging, Hillsdale, Lawrence Erlbaum. Salthouse, T.A. (1991), Mediation of adult age difference in cognition by reductions in working memory and speed of processing, Psychological Science, 2, 3, p.p. 179-183. Salthouse, T.A. (1993), Influence of working memory on adulting age differences in matrix reasoning, British Journal of Psychology, 84, 2. Scharps, M.J., Gollin, E.S. (1988), Aging and free Recall for items located in Space, Journal of Gerontology: Psychological Science, 43, p.p. 8-11. Scharps, M.J.(1998), Age-related change in visual information processing: toward a unified theory of aging and visual memory, Current Psychology, 16. Scott, L.M., R., Batra (2003), Persuasive Imagery: A Consumer response Perspective, New yersey, Laurence Erlbraun Associates.

72 ISSN 1961-7798 - Š 2009, International Marketing Trends Conference

Sundar, Shyam and Sriram Kalyanaraman (2004) Arousal, Memory and Impression-Formation Effects of Animation Speed in Web Advertising, Journal of Advertising, 33, (Spring), 7-17. Tse, A.C.B. and Lee R.P.W. (2001) Zapping Behaviour during Commercial breaks, Journal of Advertising Research, 41, 25-9. Vandewater, Elizabeth A, Victoria J. Rideout, Ellen A. Wartella, Xuan Huang, June H. Lee, and Mi-suk Shim (2007), Digital Childhood: Electronic Media and Technology Use Among Infants, Toddlers, and Preschoolers, Pediatrics, Vol. 119, No. 5, (May), 10061015


24366_int_Mise en page 1 22/12/10 12:25 Page73

JOURNAL OF MARKETING TRENDS - SUBMISSION GUIDELINES

Editorial Objectives The Journal of Marketing Trends aims to: 1. Become an academic publication which recognises the work of researchers concerned with new trends and developments in marketing 2. Encourage cross-disciplinary research on the major changes in marketing paradigms and new fields of original research in all aspects of marketing 3. Develop cross-country and multi-cultural research of the requisite level of quality which can be published in English, French, German, Italian or Spanish 4. Attract high quality authors from all continents who are developing knowledge on the major trends in marketing 5. Use the strength of the International Editorial Board to provide rapid support and feedback to reinforce the quality of the published work. The Journal of Marketing Trends is interested in publishing the following types of manuscripts derived from new perspective articles and advanced research in marketing: (a) Empirical research that report quantitative data, statistical tests and validity assessment information, (b) Exploratory research introducing qualitative data, verbatim and groundwork, (c) Theoretical research including conceptual articles, state of the art, meta-analyses, and critical essays.

Submission All manuscripts must be submitted by e-mail to Professor Andreani at: andreani@escpeurope.eu

Selection Papers are evaluated through a double blind review process which is ensured: first by the Country Editorial Board (2 anonymous reviewers); second by the Board of Editors (anonymous international reviewing process). The selection will be made on the basis of the publication rules of the major marketing research reviews, while also seeking to preserve the diversity of the various points of view. The reviewers are all marketing professors with high academic standards. Papers can be submitted in English, French, German, Italian or Spanish: first manuscript in the country language or in English. Once the country reviewing process is completed, papers are translated into english and submitted to the international reviewing process in English only. Papers will be published in two Languages: in English and in the country language, in respect of the tradition of the International Marketing Trends Conference.

Guidelines The formal publication rules are those standard for International academic publications and journals. To be published your paper must meet the following requirements: - 20 (min) to 30 pages (max) long (<30,000 characters in English - spaces not included and <30,000 characters in paper language - spaces not included)

Journal of Marketing Trends - Volume I (February 2011) 73


24366_int_Mise en page 1 22/12/10 12:25 Page74

JOURNAL OF MARKETING TRENDS - SUBMISSION GUIDELINES

- Written in English, French, German, Spanish or Italian - Manuscript file in Word format - Font Times New Roman 12 - Double spaced - Justified - Margins 2.5 cm (top, bottom, right, left) • Structure - Name of authors and co-authors on the front page - Author Note Page on the front page: present position, address, telephone number and e-mail address and any acknowledgment of financial or technical assistance - Title and content starting from the second page in English and (optional) in paper language - Immediately after the title, abstract in English (< 650 characters) and (optional) in paper language, abstract structure: main objective, methodology, results, key words in English and (optional) in paper language (max 5) - Text structure: Introduction, Conceptual Framework / Litterature Review / Research Model, Method, Findings, Discussion, Limitations and Further Research, Managerial Implications - Bibliography at the end of the article - Appendixes, Tables and Figures after the bibliography at the end of the manuscript (indicate placement in text) - Papers must contain original material neither published nor presented elsewhere prior to journal presentation with the exception of the International Marketing Trends Congress for the paper selected at the Congress • Style - The title of the manuscript: centered - Headings: > Major Headings in bold type: left-hand side, lowercase > Subheadings in lowercase, bold type and italicized (left-hand side) > Tertiary headings in lowercase, italicized and indented paragraph, also indented paragraph to start a new section > Not in quotation marks, not numbered consecutively > Headings, highlight: do not underline anything > Hard page return before and after heading - Numbers: under ten (spelled out), percentage (numerals ex. 7%), that begin sentences (spelled out), units of measure (numerals), hypotheses (numerals) - Hypotheses: bold, indented, syllabus style (H1), lower case and spelled out in text - Multiple appendixes: labelled with capital letters (A, B,...) • Tables / Figures / Mathematical Notation (academic format for formulas, figures and tables) - Tables / Figures number and title on separate lines, centered, numbered consecutively - Tables / Figures number in capital and bold type, title in lowercase and bold type - Refer to tables / figures in text by number - Tables / Figures placement: indicate placement in text (ex. insert table 3), at the end of the manuscript, follow the bibliography (first tables, second figures), one per page - Equations: centered on the page - Extensive mathematical notation: provided in a separate technical appendix

74 ISSN 1961-7798 - © 2009, International Marketing Trends Conference


24366_int_Mise en page 1 22/12/10 12:25 Page75

JOURNAL OF MARKETING TRENDS - SUBMISSION GUIDELINES

• Reference citations within the text (academic format for reference citations within the text) - Citation in the text : last name of the author and year of publication, enclosed in parentheses without punctuation for ex.“(Andreani 2008)” - The name of the author within the sentence: the year of publication in parentheses “…Andreani (2008)…” - Multiple authors: three authors (full citation); four or more (name of the first author followed by “et al.” - Series of citations: alphabetical order separated by semicolons: (Andreani 2008; Collesei 2009; Moulins 2010; Mugica 2008; Vescovi 2006; Wiedmann 2007) - Multiple citations by the same author: separated by a comma (Andreani 2005, 2009) • Bibliography and reference list style (academic format for bibliography and reference list style) - References for books: book title in italics Andreani, Jean-Claude and Umberto Collesei (2011), Marketing-Communication. Paris: Marketing Trends Association. - Reference for periodicals: name of the periodical in italics Andreani, Jean-Claude (2007),"Marketing and Diversity: the Impact of Web 2.0 Technologies", Journal of Marketing Trends, 11 (Summer), 7-14. - Reference for proceedings and edited works: lowercase Moulins, Jean-Louis (2011), “Toward a Science of Marketing Trends”, in International Marketing Trends, Congress, Vol. 9, ed. Jean-Claude Andreani and Umberto Collesei, Paris: Marketing Trends Association, 10-22. - References to Websites and URLs: lowercase Collesei, Umberto and Tiziano Vescovi (2009),"Advertisement Effectiveness: an exploratory research", (accessed February 28, 2010), [available at http://www.marketing-trendscongress.com/]. - Reference for an article in a book: lowercase Collesei, Umberto and Tiziano Vescovi (2010), "The Web Marketing: An Empirical Evaluation", in Marketing-Communication, Andreani Jean-Claude et al., eds. Paris: Marketing Trends Association, 82-99. - Newspapers: Umberto Collesei (2012), " La concurrence fera-t-elle baisser les prix du mobile ?", Le Monde, August 29, B17. - Government publication: lowercase France Ministère de la Recherche (2015), New Product Development, Paris Government Printing Office. - Multiple references by the same author in the same year: alphabetically by title 2012a, 2012b - Reference with original publication dates: lowercase Vescovi T. (1999/2012), Sociologie et marketing, Paris, PUF. - References to unpublished works: lowercase Andreani, Jean-Claude (1986), “Modelling Consumer Behaviour using Simulation Methods”, doctoral dissertation, University of Paris.

Journal of Marketing Trends - Volume I (February 2011) 75


24366_int_Mise en page 1 22/12/10 12:25 Page76

JOURNAL OF MARKETING TRENDS - NOTES

..............................................................................................................................................................................................................................................................

..............................................................................................................................................................................................................................................................

..............................................................................................................................................................................................................................................................

..............................................................................................................................................................................................................................................................

..............................................................................................................................................................................................................................................................

..............................................................................................................................................................................................................................................................

..............................................................................................................................................................................................................................................................

..............................................................................................................................................................................................................................................................

..............................................................................................................................................................................................................................................................

..............................................................................................................................................................................................................................................................

..............................................................................................................................................................................................................................................................

..............................................................................................................................................................................................................................................................

..............................................................................................................................................................................................................................................................

..............................................................................................................................................................................................................................................................

..............................................................................................................................................................................................................................................................

..............................................................................................................................................................................................................................................................

..............................................................................................................................................................................................................................................................

..............................................................................................................................................................................................................................................................

..............................................................................................................................................................................................................................................................

..............................................................................................................................................................................................................................................................

..............................................................................................................................................................................................................................................................

..............................................................................................................................................................................................................................................................

..............................................................................................................................................................................................................................................................

..............................................................................................................................................................................................................................................................

..............................................................................................................................................................................................................................................................

..............................................................................................................................................................................................................................................................

76 ISSN 1961-7798 - Š 2009, International Marketing Trends Conference


24366_int_Mise en page 1 22/12/10 12:25 Page77

JOURNAL OF MARKETING TRENDS SUBSCRIPTION ORDER I subscribe to The Journal of Marketing Trends and I want my subscription to be renewed automatically every year until expressly terminated by me.

❏ Mr. ❏ Mrs. ❏ Ms. Last name : .................................................................................................................................................................................... First name : .................................................................................................................................................................................... University/ Company : .................................................................................................................................................................................... Address : .................................................................................................................................................................................... Zip Code : ......................................................................................................City : ................................................................ Country : .................................................................................................................................................................................... Tel : .................................................................................................................................................................................... E-mail : ....................................................................................................................................................................................

SUBSCRIPTION RATE The Journal of Marketing Trends is a quarterly publication: February, May, September, November ISSN 1961-7798. The subscriptions for 4 issues start from upcoming issue.

❏ Subscription for 4 issues - Paper Printed Edition sent by mail

120 Euros

❏ Single Issue (Print) - Please indicate Issue Number ......................................................................................................................40 Euros Total: ________ Euros

❏ Payment by bank transfer upon reception of invoice ❏ Payment by bank or postal cheque to the order of “Venise-Paris Marketing” (France only) Cancellation terms - No refund on subscriptions will be made after the first issue of the journal for the subscription has been dispatched. All changes of address should be notified to the publisher at venice_paris_marketing@escpeurope.eu including both old and new address.

CONTACT INFORMATION Please return your subscription order • by mail to: Venise-Paris Marketing - c/o ESCP-Europe - 79, av de la République - 75543 Paris Cedex 11 - France - Tel : + 33 1 49 23 22 96 • or by email to: venice_paris_marketing@escpeurope.eu Venise-Paris Marketing - c/o ESCP-Europe - 79, av de la République - 75543 Paris Cedex 11 - France Tel : + 33 1 49 23 22 96 - E-mail : venice_paris_marketing@escpeurope.eu

Journal of Marketing Trends - Volume I (February 2011) 77


24366_int_Mise en page 1 22/12/10 12:25 Page78

78 ISSN 1961-7798 - Š 2009, International Marketing Trends Conference


24366_int_Mise en page 1 22/12/10 12:25 Page79

JOURNAL OF MARKETING TRENDS - SPECIAL ISSUE

CALL FOR PAPERS JOURNAL OF MARKETING TRENDS SPECIAL ISSUE ON MARKETING – COMMUNICATION TRENDS The Journal of Marketing Trends calls for a special issue on Marketing-Communication Trends. We invite papers that study marketing communication strategy and consumers and shoppers mobility related to e-media. We also invite papers that generate new theoretical and empirical insights in the area of marketingcommunication and the effects of variables, such as 360° or 365d marketing, web communication and promotion, effectiveness measurement of mix media. We invite scholars from all the disciplines within marketing to submit papers on marketing-communication trends with consumer behavior and marketing management implications to the special issue. Papers can be submitted in English, French, German, Italian, Portuguese and Spanish, and then published in two languages or in English. The submitted manuscripts should follow the JMT formats: theoretical or empirical (qualitative or quantitative) research. All submission will be subject to the peer-review process of JMT (country double blind review process + international double blind review process). We expect this special issue to conduct to significant cross-countries and multicultural fertilization and to develop the research in marketing trends. This issue is sponsored by universities involved in the International Marketing Trends Conference, as follows: Catholic University of Portugal (Portugal), Columbia University (USA), Copenhagen Business School (Denmark), ESCP Europe (France), HEC Montréal (Canada), IIM Calcutta (India), Kobe University (Japan), Leibniz Universität Hannover, MMU Business School (Manchester, UK), Philipps Universität Marburg (Germany), Poznan University (Poland), Sun Yat-Sen University(China), Universidad de Las Palmas de Gran Canaria (Spain), Universidad Pública de Navarra (Spain), Università Ca’ Foscari Venezia (Italy), Université Bordeaux IV (France), Université de Sherbrooke (Canada), University of Adelaide (Australia). Submission deadline: July 2, 2011. Expected Publication Date: September 2012 Please submit manuscripts to andreani@escpeurope.eu and designate “Special Issue on Marketing-Communication Trends”. Editors:

Jean-Claude ANDREANI, ESCP Europe Umberto COLLESEI, Università Ca’ Foscari Venezia

Co-editors: Chris HALLIBURTON, ESCP Europe (UK) Christian DUSSART, HEC Montréal (Canada) Abraham KOSHY, IIMA Ahmedabad (India) Paulo de LENCASTRE, Universidade Católica Portuguesa (Portugal) Jean-Louis MOULINS, Université Aix-Marseille II (France) José MUGICA, Universidad Pública de Navarra (Spain) Taihong LU, Sun Yat-Sen University (China) Tiziano VESCOVI, Università Ca’ Foscari Venezia (Italy) Klaus-Peter WIEDMANN, Leibniz Universität Hannover (Germany)

Advisory Board to the Special Issue: Ruth ASHFORD, MMU Business School, Manchester (UK), Prafulla AGNIHOTRI, IIM Calcutta (India), Asunción BEERLI PALACIO, Universidad de Las Palmas de Gran Canaria (Spain), François CODERRE, Université de Sherbrooke (Canada), Torben HANSEN, Copenhagen Business School (Denmark), Kemal KURTULUS, Istanbul University (Turkey), Michael LINGENFELDER, Philipps Universität Marburg (Germany) Chieko MINAMI, Kobe University (Japan), Bogna PILARCZYK, Poznan University (Poland), Pascale QUESTER, University of Adelaide (Australia), Kaouther SAIED BEN RACHED, FSEG Tunis (Tunisia), Donald SEXTON, Columbia University, New York (USA), Jean-François TRINQUECOSTE, Université Bordeaux IV (France)

There will also be ten Guests Reviewers (besides the regular JMT reviewers in this expertise area) handling the special issue manuscripts.

Journal of Marketing Trends - Volume I (February 2011) 79


24366_int_Mise en page 1 22/12/10 12:25 Page80

JOURNAL OF MARKETING TRENDS - NOTES

..............................................................................................................................................................................................................................................................

..............................................................................................................................................................................................................................................................

..............................................................................................................................................................................................................................................................

..............................................................................................................................................................................................................................................................

..............................................................................................................................................................................................................................................................

..............................................................................................................................................................................................................................................................

..............................................................................................................................................................................................................................................................

..............................................................................................................................................................................................................................................................

..............................................................................................................................................................................................................................................................

..............................................................................................................................................................................................................................................................

..............................................................................................................................................................................................................................................................

..............................................................................................................................................................................................................................................................

..............................................................................................................................................................................................................................................................

..............................................................................................................................................................................................................................................................

..............................................................................................................................................................................................................................................................

..............................................................................................................................................................................................................................................................

..............................................................................................................................................................................................................................................................

..............................................................................................................................................................................................................................................................

..............................................................................................................................................................................................................................................................

..............................................................................................................................................................................................................................................................

..............................................................................................................................................................................................................................................................

..............................................................................................................................................................................................................................................................

..............................................................................................................................................................................................................................................................

.............................................................................................................................................................................................................

.............................................................................................................................................................................................................

Imprimé en France par EMD S.A.S. N° imprimeur : 24366 - Dépôt légal : décemb re 2010

.............................................................................................................................................................................................................

80 ISSN 1961-7798 - © 2009, International Marketing Trends Conference

JMT  

kjlkjlkjljd

Read more
Read more
Similar to
Popular now
Just for you