The article’s premise is that operational skill no longer confers advantage. Leaner manufacturing, higher quality and better distribution have become table stakes, and the new source of advantage is customer centricity.
Data alone does not deliver that. What separates winners is the ability to turn data into insight about consumer motivation and to turn that insight into strategy, which the authors call the insights engine.
Key findings
- The Insights2020 research behind the article involved in-depth interviews with more than three hundred and fifty business, marketing and insights leaders, plus online surveys of more than ten thousand practitioners in sixty countries.
- Of the three categories of customer-centric organization the research identified, the insights engine had the largest impact on revenue growth.
- Overperformers are twice as likely to have insights leaders reporting to a senior executive, at twenty-nine percent against twelve.
- The gaps are wide across every characteristic measured, including linking disparate data sources at sixty-seven percent against thirty-four, and whole-brain thinking at seventy-one percent against forty-two.
- Seventy-nine percent of insights functions at overperforming companies participate in strategic decision making at all levels, against forty-seven percent at underperformers.
From service unit to strategic voice
The article opens with a moment that demonstrates the shift. When Unilever announced an initiative to move resources to local markets, the analysis behind it came from a presentation by the head of the insights group, which focused attention, catalyzed the strategy conversation, and led to changes in organization and mindset.
The authors are explicit that this would once have been unthinkable. A decade earlier, strategic involvement of this kind by a customer intelligence operation was almost unheard of, and the market research department was typically a reactive service unit fielding requests and producing reports.
The stated mission of Unilever’s group deliberately omits the word insight, because insight is a means to an end. The end is action that drives business growth.
Synthesizing data into one version of the truth
What used to take months and cost millions can now be done in days for a fraction of the price, so the advantage no longer lies in the quantity of data a firm can amass but in its ability to connect the dots.
The first challenge is integration. Structured and unstructured data from sales figures, media spending, call center records and social monitoring is customarily owned by different teams, and may run to tens of millions of pieces.
Working with the technology function, the group built a global marketing information system that presents data in consistent formats so that everyone sees the same numbers, derived the same way. It calls this one version of the truth, and it removed most of the arguments about definitions and methodology.
The worked example is cholesterol. Research showed that behavioral change only sticks if people use the products for at least three weeks, and that the way to get that commitment is through peer pressure. That produced a program challenging a whole town to lower its cholesterol, now running in more than ten countries, with eighty-five percent of participants lowering theirs.
Independence, planning, collaboration
Superior insights groups sit outside marketing and report into the C-suite, which lets them stay objective, collaborate as equals, and challenge or even set direction. At Unilever the group reports to an executive board member and has autonomy over its own budget.
Independence is what made advertising pretesting possible. Using consumer surveys and software that reads facial expressions, weak advertisements are killed before airing. Creators initially saw the program as a threat to creativity and resisted it, and now embrace it, partly because successful advertisements count toward their bonuses.
Integration into the planning cycle is what turns advice into decisions. Substantially more overperformers than underperformers include insights leaders at all key stages of the cycle, at sixty-one percent against forty-six.
Two custom tools support the two planning questions. One mines millions of data points on consumer demand to quantify the value of deeper penetration, answering where to play. The other gives a single-screen view of a brand’s performance against its category, answering how to win.
Collaboration is structural rather than cultural. The insights teams are colocated with the presidents of the product categories they cover, and being held accountable for business results gives the group a reason to work with every commercially oriented team.
Everyone as a source of insight
The company treats insight as available to anyone at any time. Through one program, employees from factory workers to the chief executive connect directly with customers, and about thirty thousand people take part annually.
An always-on platform lets any employee arrange a video conversation with a consumer anywhere, and observations are captured in an in-house app, stored centrally, and analyzed with video mining to find behavioral patterns across regions.
The results are specific. Reports from visits to kitchens in China revealed that high heat and tight space make grease buildup a common problem, and brand teams began working on what product innovation could address it.
Experimentation and looking forward
Overperformers are three times as likely to embrace a culture of experimentation, at forty percent against thirteen. Unilever formalized this through an incubator that runs hackathons, a collaboration platform, and challenges posted as requests for proposals.
One initiative has start-ups pitch new technologies to an executive team in five minutes with five minutes of questions, after which the team votes on what to pilot. In two years it screened more than six hundred and fifty technologies, piloted more than a hundred and seventy-five, and scaled up thirty-seven.
One of those start-ups uses wearable technology to observe what people actually do rather than what they say they do. Test subjects wear a headset during routine tasks, and the analysis surfaces behavior they are unaware of, such as the living room taking longer to clean than the bathroom rather than the reverse.
The forward-looking work goes past monitoring the present. A program built with technology partners analyzes around a billion hair-related searches a month to spot styling trends, then rapidly creates how-to videos featuring products without directly promoting them. It has had more than a hundred and twenty-five million views and is three times as likely to drive purchase as conventional advertising.
A dedicated team studies the macro forces shaping the world, including economic and technological growth shifting east and south, and growing environmental stress. One such conversation led a brand team to data showing that many deaths of children under five could be prevented by handwashing, producing an education program that changed the behavior of three hundred and thirty-seven million people in twenty-eight countries.
The people characteristics
The first is whole-brain thinking. Insights groups historically focused on analytics, and that left-brain orientation served them well, but the authors argue today’s teams must think holistically and exercise creative skills too. Achieving the balance takes both recruiting and changing the existing organization.
Workshops push people out of their default thinking styles and into creative problem solving with colleagues they would not normally work with. One brought marketing, research and development and insights people together on conditioner sales in Southeast Asia, and the insight that consumers would not risk buying a product whose benefit they were unsure of led to an inexpensive trial-size packet.
The second is business focus. Seventy-five percent of respondents at high-performing firms saw their insights function as business-focused against fifty percent at low performers, and Unilever links staff bonuses to wider business unit performance to create shared accountability.
The third is storytelling. Sixty-one percent at overperformers agreed their insights people were skilled at conveying messages through engaging narratives, against thirty-seven percent. The group moved away from charts and tables toward provocative storytelling, on an ethos of show, don’t tell.
The most literal example is an initiative on older consumers. Rather than reporting how they struggle, marketing executives wore old-age simulation equipment and tried to read labels and handle products. One outcome was ice cream packaging that is easier to read.
Scope and limitations
The research divides respondents into two groups by three-year revenue growth relative to competitors, so the findings are correlations between self-reported practice and self-reported performance. The authors are also candid that even the most advanced insights engine cannot make a firm customer-centric on its own, which requires leadership from the top.
Source
This page summarises Building an Insights Engine: How Unilever got to know its customers, by Frank van den Driest, Stan Sthanunathan, Keith Weed, September 2016.