One little white lie at a time, some of the world’s biggest businesses are being undone by bad consumer data, and they don’t know it yet. Now that new tools like synthetic audiences are being thrown into the mix, the boardroom faces a novel problem.
Let’s start at the beginning. It doesn’t take a genius to know that businesses need good data to understand their consumers. To market effectively, brands need to know what their audiences are thinking, feeling and doing.
For almost 20 years, I’ve advised brands on how to and how not to connect with young audiences through “culture.” Working at VICE at the height of its influence, our job when we walked into the boardroom was essentially to scare the living daylights out of senior marketing execs, who had no idea about the attitudes, behaviors, tastes and desires that defined the young people they were trying to talk to. Working with us was the only way they would save themselves from certain death.
Millennials were shaped by the pillars of their parents’ realities crumbling around them. That this would in fact become the mood music for the next two decades was at the time unknown. But the lack of accurate data available to decision-makers to educate themselves was itself what provided the opportunity for VICE to flourish.
Fast forward 15 years and much remains the same. Big brands are drowning in lines of consumer data they have little use for, and their competitors are acting on the same generic narratives as them. Legacy polling companies have always told brands X percent of people think Y about their product. They ask direct, generic questions and they get direct, generic answers. These answers become polemic (for want of anything more interesting or, dare to say it, insightful) and the world keeps on spinning.
It might seem like a small, or even interpersonal, problem. But brands regularly build multimillion-dollar projects on the backs of the lies people tell about themselves. Serious money is going to the wrong places: a study last year by the research organization Esomar estimated the “insights industry” was worth $150 billion, with $56 billion spent on market research alone. How many of the people holding these budgets would tell you that they are getting full value for their money (and have accurate data to back this up)?
My co-founder, Alfred Malmros, is among those stung. He ran a startup called Anyone, an app that provided five-minute advice calls from anyone in the world. Like all responsible founders, he wanted to figure out whether people wanted what he was building and launched a product-market-fit survey. The results filled him with confidence that everyone wanted it.
That prediction didn’t wash with the download statistics, which peaked at 25,000 and dwindled to 1,000 active users over time. The same people who claimed they would love an on-call advice service never downloaded the app. It was this painful lesson that taught us that the wrong questions generate the wrong answers.
If my co-founder had known at the time why we lie and how lies fuel bad data, the crisis could have been averted.
Why we lie
Picture this. You’re at a dinner party with a half dozen people you don’t know that well. The conversation is about climate change and the scale of the challenge it presents to humankind. Everyone concurs: It’s a big one.
A guest has just finished saying that climate change keeps her up at night, to thoughtful nods from the rest of the table, and then she turns to you and says: What do you think?
In the moment most of us would agree. It would take a contrarian to take a different or contradictory position. Or even to admit that, honestly, what actually kept us up at night was our mortgage or the dreaded monthly 1:1 with our manager the next day.
There are countless examples of this with brands: Ask people why they want an expensive watch, and they will say the craftsmanship; ask them why others want one, and they will say status.
Whether conscious or not, we so often default to dishonesty. We lie to make sense of ourselves to others, to make sense of a world that is increasingly nonsensical. It’s not a moral failing, but it is a sticky conundrum for the businesses trying to understand us.
So how can businesses ask better questions?
The first solution is simple: Ask people what their friends think. Academics have shown that when people answer questions on behalf of their social circle instead of themselves, they effectively report on many people beyond themselves, which tells us a lot more about society than a direct question to one person.
Second, stop asking people what they think only in that moment, and broaden it to what they’ve already experienced and what they expect to experience. Ask the same question about five years ago, today and three years from now and you’ll get back data on the direction of travel in beliefs and behavior, not just a snapshot.
My business partner and I built our research consultancy on these parameters. Consumer data sets are so often made up of people’s projections and manifestations, the ways in which they’d like to be seen. To get to real truths, even better, hidden truths, researchers must solve for this.
Even better, design for it. When enough intelligence (human or artificial, ideally a blend of both) goes into crafting a question, the question itself is enough to unlock new futures and opportunities, and for those who ask the question to see their worlds anew.
Why should this be keeping C-suite leaders awake at night?
Because bad data alone isn’t a new problem. It’s been compounded by the rise of synthetic audiences: AI-generated simulations of real consumers that businesses can interview and test ideas on before launching them.
In theory it’s useful, but when these audiences are trained on flawed consumer data, the consequences are catastrophic.
If businesses aren’t paying attention to the nature of the data going into these models, the results may look good, but they will be fundamentally inaccurate. It’s a problem we’re already seeing play out. Taking a broken methodology and feeding it bad data will only make the wrong answers arrive faster.
With or without AI acceleration, if more businesses knew how to ask consumers the right questions, we would break the cycle of bad data and bad decisions that has existed for decades, averting C-suite crises. The companies that continue to solve for an imagined version of their consumer will lose to the companies solving for the real one.
The opinions expressed in Fortune.com commentary pieces are solely the views of their authors and do not necessarily reflect the opinions and beliefs of Fortune.
