• by David Stanton , Op-Ed Contributor, May 6, 2025

Is artificial intelligence (AI) the solution to all of our consumer insight problems?

Is it a semi-reliable “direct report” that can just take on a few basic tasks? 'Or is AI a merchant in “snake oil” that should be approached with caution -- perhaps avoided completely?

All three points of view were on display last week at the annual IIEX (Insight Innovation Exchange) event, held in Washington D.C. for the first time this year.

A few years ago, AI was just a blip on the research horizon. But at this year’s Greenbook-hosted event, some presenters felt a need to apologize if they did not focus on AI -- and sessions with those two letters in their titles invariably drew bigger crowds.

In my decidedly non-representative sample of presentations and conversations, IIEX 2025 -- an event that was well organized, comprehensive, and congenial -- seemed to show an industry at a crossroads. 

The fact is that human-answered surveys are alive and well, and are still relied on to collect insights that guide many of the world’s best-known brands.

Some clients are coveting human points of view as never before. (More on this later.) But AI is also being widely used and sometimes extravagantly hyped -- to analyze data, prepare questionnaires and summaries, and even conduct full-blown AI-only focus groups, moderator included.

To be sure, AI is prompting a reckoning -- if not an earthquake -- in almost every marketplace. But because AI technology can actually invent content (including survey answers), it can deliver an extra dose of disruption to consumer insights. 

The IIEX event brought together insights leaders from a host of both traditional and e-world brands -- eBay, Pernod Ricard, CapitalOne, Mars Food, Estee Lauder, Kraft Heinz, and many more.

Snapchat’s Aarti Bhaskaran and Amazon’s Caryn Akons took part in two sessions each, generously sharing their insights, goals, and challenges.    

This year’s IIEX often echoed the headlines of the day. On kickoff morning, The New York Times blared: “US Economy Shrank in First Quarter, in Reading Clouded by Messy Trade Data.”

And the need to work within shrinking budgets and save both time and money will likely drive many clients and suppliers in the direction of AI -- just as they flocked to Internet panels and surveys twenty-some years ago.

Meanwhile, the market research trades were full of the shocking (but maybe not so so shocking?) news of three research firms charged with survey fraud -- possibly robbing clients of millions in fees, and creating the potential for mistaken client decisions.

As Steve Male, executive vice president of innovation and strategic partnerships at The Logit Group, observed: “This was not an off-shore fly-by-night – they sold on data quality.”

Speaking on the same panel as Male, Procter & Gamble insights executive Tia Maurer said flatly: “The ecosystem is broken … We are all complicit in this -- all of these companies have chased the cheaper price.”

When MR suppliers and clients began to opt for faster and cheaper Internet-based research decades ago, talk about data quality became more scarce. But now AI bots with seemingly unlimited capacity have been added to “hyper-responders” taking 20 or more surveys a day to create a major trust problem for the insights market.

According to Carrie Campbell of Case4Quality -- a brand-led coalition focused on research quality -- every major data provider is producing at least some fraudulent data.

“The foundation of what we have has been called into question,” Quest Mindshare Research Director Scott Worthge said plainly, with data-quality issues “tearing up the landscape. There are going to be companies that need to go out of business.”

Meanwhile, on a futuristic plateau overlooking the exhibition floor -- like a spaceship landing on the conference -- major event sponsor Qualtrics held presentations extolling the benefits of “synthetic” data, which is created exclusively by AI.

Leveraging algorithms trained on the preferences and mindsets of consumers, synthetic data can deliver results in minutes rather than weeks, at a fraction of the cost of human-generated studies and with PI intact.

The vast reach of AI into every aspect of research also prompted another conference theme -- the demand for insights from verified and verifiable humans. Multiple conversations and presentations at IIEX pointed to an absolute need to know, beyond a shadow of a doubt, that data is human generated --  because humans are still the customers everyone is selling to.

Maurer reported that she has at times opted for qualitative (focus group) research over quantitative to be sure her respondents are real people -- because, in her words, “I don’t want to spend a week cleaning my data.”

There is a clear strain of wariness in some attitudes to AI. More than a few IIEX presenters -- in some form or other -- assured audiences that AI was merely going to “supplement” the work of humans, not replace it.

Meanwhile, newer vendors like ONE Strategy promise to take clients “from brief to breakthrough in hours,” with an all-AI approach that guarantees faster, better, and cheaper -- and that has won major brand clients, from Philips to Unilever.

The truth is that many of the insights industry’s stalwarts are simply laying AI onto existing products and practices -- and the new technology does allow them to deliver analyses, reports, and other services at record speeds.

But concerns about AI “hallucinations” and other mistakes persist -- and these traditional researchers are not ready to let go completely to the AI’s abilities.

Meanwhile, more than one long-term industry observer I spoke to questioned whether clients are truly that concerned about data quality today, with so many other issues to deal with -- squeezed budgets, shrinking teams, and more.

And with research fraud now officially a topic of open concern, what is to keep buyers from leaning fully into AI?

At odd moments during the event, tracks from Billy Joel’s “The Stranger” wafted through the venue sound system. With their wistfulness for lost innocence, Joel’s songs seemed to provide a kind of soundtrack for the conference. “Then the King and the Queen went back to the green,” Joel sang, “but you can never go back there again.”

With no option to go back, where will the insights industry head from here -- with AI playing the roles of superhero and supervillain alike in a breakneck, high-stakes drama?

Let’s hope that a fresh, shared vision for maintaining quality, embracing change, and delivering value comes into focus soon.