The Data-Empowered Organization

Page 1

MANAGEMENTThe Magazine of the Rotman School of Management UNIVERSITY OF TORONTO FALL 2022 IS YOUR FUTURE-READY?ORGANIZATION PAGES 18, 42, 60, 78ThatValueBuildingLasts PAGE 48 How Navigateto Metaversethe PAGE 30 ofandLeadershiptheIllusionControl PAGE 24

The Future-ProofIssue

Tell You

The OrganizationData-Empowered

A FEW YEARS BACK, the media was full of ar ticles about how Big Data would solve a per rennial challenge: gaining valuable customer insights. Today, it is everywhere because of the growth of devices recording data and the connectivity between those devices through the Internet. The problem is this: Big Data systems can in deed find subtle patterns in data sets, but they can’t tell us which correlations are truly meaningful. As we know, corre lation does not equal causation. However, we’re hardwired to look for patterns, and when we see lines sloping together in our data, it’s hard for us not to believe there’s a causal re lationship.

98 / Rotman Management Fall 2022

The fact is, when organizations gather customer data without rhyme or reason, without clearly defining the problem that needs to be solved, no software or analy sis tool will generate beneficial insights. Recent research has found that as data sets grow larger, they contain sig nificantly more arbitrary correlations. These correlations were proven to appear solely due to the number of items in the data sets.

In statistics, this is known as a ‘spurious correlation’ a mathematical relationship where two or more items or events are associated but not causally related. It can be due to either coincidence or the presence of another factor (seen or unseen). In the case of increased ice cream sales and shark attacks, that other factor is temperature: Higher temperatures cause more of us to buy ice cream, as well as go to the beach for a swim.

Tyler Vigen, the author of Spurious Correlations, high lights this issue through a series of comical-yet-technicallycorrect line graphs showing how monthly U.S. ice cream sales track in perfect parallel with forest fires and shark attacks. Simply because ice cream sells more in the summer, there’s no real correlation with an increase of shark attacks during the same period, but the data can cause us to perceive it.

Put simply, more data does not necessarily equal more useful information.

POINT OF VIEW Andrea Belk-Olson, Author, What to Ask: How to Learn What Customers Need But Don’t

Adding Data versus Adding Information

rotmanmagazine.ca / 99

In the not-so-distant past, computers were only given to ex ecutives or engineers. But today, everyone’s got one, along with the essential tools and training to use it effectively. The same should go for data. Creating a data-empowered orga nization requires restructuring work processes and behav iours to enable employees to use data to make better busi ness Adecisions.perfectexample of this is 7-Eleven Japan, where op eration field counselors (OFC) visit each of the company’s 21,000 stores twice a week, helping salesclerks learn to use

Customer retention rates with Millennials have declined, creating a loss of $465k in revenue this year alone, which may be due to three new competitive products in the market (why), creating the need to examine how customers perceive and utilize our current offerings (wherefore).

Wherefore Statement defines not only a problem to be solved and its perceived cause but also how you will go about trying to solve it. This approach shifts the perspective from ‘things to do’ to ‘problems to solve.’ For ex ample, a traditional goal and set of objectives might take the following form:

Goal: Grow customer retention by 25 per cent in the millennial market.

Before gathering any new data, companies need to first understand their objectives.

An objective sets a specific and focused strategy to at tain a particular goal. Bad objectives read like a to-do list. Most people create objectives as a target or aim rather than a strategic approach to solving organizational challenges. For example, an objective might be ‘creating a customer loyalty program’ for the goal of growing customer retention. There is no clarity as to why this approach is preferred over others and no specific parameters for determining success or fail ure (other than creating the program itself).

John Kelly, IBM’s father of Watson, once said that 80 per cent of data is untouched — meaning it’s never actually used to make improvements or changes to the customer experience. No matter how much data or information you have on your customers, it is useless unless it is translated into action — something executable that can be tested, mea sured and refined. And the more we enrich our data with meaning and context, the more knowledge and insights we will get out of it, enabling better-informed and data-driven decisions.

While these objectives may have merit, they are a shot in the dark, because they are solutions to an undefined prob lem that is impacting customer retention. Has retention been declining only in this segment, or in others, as well? What is

Back to Business 101

Executives are infatuated with goals. The prevailing belief is that to grow an organization, you need specific, measurable, actionable, relevant and time-bound goals. Yet they usually manifest as broad, generic and unrooted in research. Things like ‘ensure success for enterprise customers’ or ‘increase sales leads by 10 per cent.’ When you establish bad goals, you get bad objectives.

A Why and Wherefore Statement is a situation-andapproach model, highlighting the information we know to day, our assumptions about the cause and our approach to validating or invalidating those assumptions. Using the ex ample above, a Why and Wherefore Statement might take this form:

Enabling Data Dissection

Before considering gathering new data, companies need to first understand their objectives, then determine the type of information needed to address them.

What sets successful organizations apart is understand ing how their goals can be achieved. To truly understand and define the how, you must first understand the what. And I propose that the missing component is a Why and Where fore Statement.AWhyand

Objectives: Conduct a video-marketing campaign for at-risk customers; create a rewards program; develop a cus tomer advisory board.

a healthy retention rate for our industry and offerings? What outside factors have been impacting retention rates, such as the entrance of new competitors or technologies?

Organizational goals are empty without a comprehen sive understanding of the issues underlying them. A goal of growing customer retention by 25 per cent might seem ad mirable, but without more context, there’s no way to deter mine whether it’s worth pursuing.

Healthcare and life sciences meet wwww.rotman.utoronto.ca/GEMBA-HLSLeadersbusiness.wanted.

rotmanmagazine.ca / 101

Many companies talk a good game about being data driven, but when data is only accessible and understood by a select few, the effectiveness of an insight is inherently limit ed. By committing to an immersive and pervasive approach to using data across the entire company, you can capitalize on its value over the long term.

Organizations don’t need to just learn how to use cus tomer data, but also how to look beyond the numbers. A great example of this skill in action is Netflix. The idea of giving viewers a digital catalogue for streaming online not only killed video rentals, it also made Netflix a cultural phenomenon, as viewers embraced a new way of getting their entertainment. In the very early days, Netflix didn’t have original content, but they were obsessed with customer data and, more im portantly, customer behaviour. They found that 76 per cent of viewers said ‘watching multiple episodes of a great TV show’ was a welcome refuge from their busy lives. They also found that the tipping point at which viewers actually got hooked on a show was never the pilot episode, but a few episodes further in.

The lesson I hope you will take away is this: never pass up the chance to get more context about a problem. The more informed you are, the better your assumptions and data interpretations will be. Remember, data doesn’t auto matically provide us with answers; it’s how we interpret it that creates aha moments.

Harold Geneen, author of the 1984 book Managing, once said, “When you have mastered numbers, you will in fact no longer be reading numbers — any more than you read words when reading a book — you will be reading mean ings.” Empowering your entire organization to understand, evaluate and embrace customer data requires commitment. That means investing time and resources today to reap the benefits over the long term.

Andrea Belk-Olson is the author of What to Ask: How to Learn What Customers Need but Don’t Tell You (Benbella, 2022). She is also a Lead Instructor for the University of Iowa Venture School and CEO of Pragmadik, a Behavioural Science consultancy.

Netflix capitalized on these behavioural insights by de veloping a strategy to release all episodes of shows simulta neously, instead of releasing them one at a time as had been done since the dawn of television. The term ‘binge watch’ became the Collins English Dictionary word of the year in 2015. Their success came down to not only having data, but also applying knowledge gained from that information. Since Netflix was able to look beyond the numbers, it was better equipped to address what customers were wanting. While some companies can draw valuable knowledge from data, others can be misled. An energy company ob served customer attrition correlated to the frequency of customers’ Google searches for an energy supplier. They deduced from the numbers that online searches were re

sponsible for 65 per cent of churn. Yet when they examined customer behaviour, it revealed the decision to switch pro viders was triggered by something else: extensive competi tor advertising and promotional pricing. Google searches were not the cause of attrition, because people had already made up their minds by the time they started searching.

The fact is, numbers can never tell us the whole story. In the 1960s, a group of economists did simple extrapola tions of past costs in an attempt to predict the long-term program costs of Medicare. Their forecast was off by a fac tor of more than 1,000 per cent (!), with actual costs being less than 10 per cent of what was projected. Why? Because human factors weren’t considered. The new Medicare pro gram included financial incentives encouraging innova tion: New medicines were invented that reduced treatment costs, and new treatments were developed as lower-cost remedies to specific health conditions. This is why the fi nancial experts designed such an incorrect forecast. They oversimplified the context to get easy figures. And in doing so, they chose not to consider the contextual and behav ioural effects of the people involved.

data effectively. They compare each clerk’s hypotheses on what would sell during the prior week versus what actually sold. They discuss how to analyze customer data and how to improve performance for the next week. The OFC is a fulltime position to which high-performing salesclerks can be promoted.

This same pitfall can occur with pricing strategies. Johnson Instruments produced a unique safety device for operators of heavy equipment that would avert physi cal damage to the user — and major financial liability to the purchaser. For some reason, it wasn’t selling, even though it was of higher quality and lower price than others on the market. They later realized it wasn’t selling because it was priced significantly lower: the price made people perceive it as a low-quality offering, and therefore a high-risk purchase in an already high-risk business. They raised the price 30 per cent, and eureka: sales volume increased 46-fold.

MANAGEMENTThe Magazine of the Rotman School of Management UNIVERSITY OF TORONTO HOW AI WILL AFFECT BUSINESS PAGES 44, 62 ChangeArtTheof The Origins of the Gender Gap PAGE 38 In defence TroublemakersofBehaviouralHarnessing Insights ApproachesBehaviouraltoDiversity PAGE 26 The OrganizationFearlessTheSeedsofChange PAGE 80 MANAGEMENTThe Magazine of the Rotman School of Management SPRING 2021 IssueHealth(y)The FIXING A BROKEN SYSTEM An forRecoveryEconomicPlanCanadaCrisis C-SuiteLessonsManagement:fromtheLeading the Way to CDL’sRecovery:Moonshot(s) Explore more rotmanmagazine.ca

Turn static files into dynamic content formats.

Create a flipbook
Issuu converts static files into: digital portfolios, online yearbooks, online catalogs, digital photo albums and more. Sign up and create your flipbook.