Modern Marketers and the Lost Treasure of Psychology-Based Segmentation

A 2002 survey found the practice of customer segmentation in precipitous decline among brands, to such a degree that the marketing world is only just now beginning to recover. What’s next for consumer understanding?

As long as people have peddled products, they’ve bemoaned not knowing enough about buyers. There’s evidence that Bronze Age Greeks faced a customer research dilemma that’d be familiar to modern marketers: Your Cypriot, Levantine and Minoan neighbors each have distinct tastes in trade goods, but aren’t close enough for you to pop over and ask. How do you plan one shipment (literally, a ship) that delivers to all what each most wants?

The acute need for insights is what inspired the use of quantitative questionnaires in the 1920s, psychoanalytic focus groups in the 1940s, and a precise combination of the two in the 1980s through 1990s when big agencies such as Ogilvy & Mather, Saatchi & Saatchi and WPP achieved what marketing professor Angus Jenkinson has called “the gold standard” for customer segmentation—a combination of quantitative analysis and sweeping ethnographic studies that gave rise to iconic campaigns like Got Milk? and the Energizer Bunny. Then, the whole field quietly fell to pieces.

In 2002, a survey found the practice of customer segmentation in precipitous decline among brands. Only one in five companies were still carrying out the basic statistical analysis and just over half were no longer recognizing consumer behavior in their planning. What was responsible? A lethal combination of algorithms and the internet. Together, they impoverished the field of customer understanding to such a degree that the marketing world is only just now beginning to recover.

The Desertification of Customer Analysis

Even during the earliest days of mass marketing, it was understood that you had to know your customer personally. “The consumer is not a moron. She is your wife,” David Ogilvy famously quipped. Brand managers gathered focus groups and observed customers in their own homes. Advertisers unleashed hordes of surveyors to quiz pedestrians about the ads they could recall, and marketers borrowed heavily from the sciences.

Many of the marketing titans of the latter half of the 19th century got their start in the U.S. government’s WWII propaganda arm. This included Edward Bernays, author of Propaganda, whose work borrowed heavily from the ideas of his uncle—a one Sigmund Freud. Larger than life agencies attracted top minds straight from Ivy League schools, and ethnographers, sociologists, anthropologists and philosophers flooded their halls, applying scientific rigor to understanding not just what consumers did, but the thinking that motivated them to do it.

Then the internet arrived.

Suddenly, vast tranches of customer information became visible online and available to analysts, and companies could use computers to run massive calculations relatively cheaply. Tech companies spread the gospel of machine power as a tool for surpassing the surly bonds of human understanding and a deep skepticism spread about whether fat-fingered and error-prone hominids even had a place in the room. The success of algorithm-driven companies such as Google, Yahoo!, Facebook and YouTube made humanity appear on the verge of fulfilling the 19th century mathematician Pierre-Simon Laplace’s postulate that with enough information, humankind could predict anything.

At the same time, the internet was changing how information was distributed. Marketers looked more to Google search than they did the venerable torchbearers of knowledge, universities and government, and the blogging revolution made non-experts the loudest voices in the room. One might liken it to a dark age where the wisdom of the ancients—the gold standard for segmentation—was forgotten. Citizen-experts began a reign of entropic chaos that reduced the field of targeting to soundbites and self-serving articles written by content managers more concerned with selling subscriptions than upholding scientific rigor.

And that brings us to today. Marketers look at their data more than ever but use faulty methods and rely on personas that are mostly stereotypes and fall short of the golden age’s standards. Forty-two percent of marketers do not even talk to customers when building customer segmentations, reports the Content Marketing Institute. Last year, the Wharton School of the University of Pennsylvania found 59% of marketers guilty of what’s known as “p-hacking,” or erroneously stopping A/B tests early when they get the result they want. An entire generation of marketers has grown up believing in ‘personas’ as analogous with ‘psychographics,’ an unscientific term of art popularized by the ratings company Nielsen to help TV advertisers sell space. It is de rigueur to append a random stock photo to a catchy, alliterative name and applaud it as insight. An emaciated shadow of their former selves, most marketing organizations slouch toward Googles or Facebook’s advertising tools and bid the algorithms to sort it out for them.

Rome fell to barbarians in 476 C.E. The city burned and the wisdom of the Greeks was lost to the Western world—but only for a time. The writing of Plato and Pythagoras survived in manuscripts and scrolls stored in Egypt and Syria and historians credit their reintroduction with the great flowering of philosophy and thought that we know as the Renaissance. So too, one hopes it may turn out, with customer segmentation, for all the knowledge of the gold standard of segmentation has been preserved within the tower of academia. Now, the people who have nurtured that flame—sociologists, psychologists and anthropologists—are making a comeback on the marketing stage.

A Segmentation Renaissance

In summer 2016, the Hilton Hotels marketing team got hung up on a question. They were puzzled why their hypotheses for programmatic ads were so often wrong. They understood the demographics of those to whom they were advertising, but they didn’t understand why people purchased or what was motivating them. So they did an unusual thing and approached a team of University of Cambridge psychologists who turned them on to psychometrics—a method for quantifying the more subjective, but arguably more important, aspects of purchase journeys: consumer attitudes and interests. The Hilton team conducted surveys, produced personality data and used it to run an ad that outperformed the control group by 340%.

Word has spread. Marketers increasingly solicit advice from octogenarian organizations like the Psychometric Society, founded in 1935, and consulting firms like Deloitte are dredging up old papers to release new research on the value of accounting for emotions in marketing. “Talking to friends in the private sector, it was obvious that an understanding of individual psychological differences had been lost,” says Avi Tuschman, a Stanford alumnus and internationally recognized expert in heritable psychometric traits, who in 2015 founded a startup called Pinpoint Predictive that offers psychometric AI-powered insights and predictive personalization to marketers. “It was just brute force. The delta between what marketers are doing and what’s available in academia is just so much bigger than people realize. But now I think now they’re starting to.”

What’s particularly interesting is that this newest wave of psychometric data startups like Pinpoint are taking marketers somewhere they haven’t been before, marrying new and old: personality insight with machine learning algorithms. Rather than conduct costly, sweeping ethnographic studies, Pinpoint uses data science to extrapolate a truth set for a portion of U.S. adults across the entire population. This lets it offer personality insights as service, based on as little information as a customer’s email. Services like this may soon democratize personality-based segmentation science for even small business owners and help all marketers recover a treasure that’s eluded them for decades: the power to message people as the unique, multi-faceted individuals that they are.

“The ability to leverage emotionally intelligent platforms to recognize and use emotional data at scale is one of the biggest, most important opportunities for companies going forward,” wrote Deloitte in a 2019 report on emotions in marketing. And so it may be: Emotions account for at least 95% of our decision-making. Our inherent traits such as intellect, modesty, self-efficacy, vulnerability and assertiveness make us who we are and invisibly influence our every action, and these aren’t captured in the demographics most marketers use, or available outside the walled gardens of the advertising duopoly.

For the modern marketer who is in tune with these delicate affectations in their customers, the marketing world may be as intelligible as the night sky to an ancient mariner, or the world of radio advertising to an ’80s ad exec, and then some.

Photo by by Ramón Salinero on Unsplash.