AI is ‘vanillizing’ insights, diluting market research's edge: Nikolai Reynolds, Ipsos

The market research industry is at a critical juncture. As it rushes to scale more rapidly and cut costs through artificial intelligence, a paradox emerges, the very tools designed to sharpen insights may be flattening them. Nikolai Reynolds, Global Head of Product Testing at Ipsos, describes this as the “vanillization” of insights, a growing concern as AI-generated content continues to feed back into AI models, creating increasingly generic, context blind outputs that lack human understanding.

“Repeated training on synthetic or AI-generated data dulls the uniqueness of insights”, warns Reynolds. He argues that fresh, real-world data, rooted in cultural nuance and human behaviour, is essential — especially in an era where globalisation is giving way to more fragmented, tribal markets. Reynolds underscores that Ipsos’s solution is hybrid thinking, or HI + AI, which combines human truth with artificial speed.

This is a pivotal moment as the industry embraces AI: will we allow computers to shape insight in their image, or will we maintain human understanding at its core?

AI Shift from Hype to Hygiene

Since the launch of GenAI, the pace of transformation in market research has accelerated dramatically and Ipsos has tried to embrace this change. A prime example is its proprietary AI platform, Ipsos Facto, now used by 80 percent of its global client service teams. This secure platform houses multiple large language models like Gemini, ChatGPT, Claude and other such platforms, supporting both operational work and strategic innovation.

Reynolds explains how clients are not just expecting, but actively demanding, AI integration. One large global brand now mandates that 50 percent of its projects be AI powered, being clear evidence that AI is no longer optional.

“You can now generate a proposition directly from a product test, or design packaging before it physically exists,” he shares, noting the collapse of traditional silos between R&D, marketing, and product development.

However, Reynolds emphasises that without human validation, the value of AI diminishes. Flashy outputs are not enough; they must be tested for actual impact on consumer behaviour.

HI + AI: Context Over Convenience

As artificial intelligence continues to take over the research landscape, Ipsos is making a deliberate case for balance and not replacement. The heart of its strategy is “HI + AI”, a methodology that blends the speed and scale of AI with the depth and understanding of human intelligence. This means not just analysing data, but comprehending consumer behaviour through emotions, gestures, local culture, and lived context.

“You can’t rely entirely on probabilistic models to predict how a concept will perform in, say, Delhi or Mumbai,” says Reynolds. “Context is king.” To further reinforce this sentiment, Krishnendu Dutta, Practice Lead for Innovation and Market Strategy, says, “We don’t believe in sacrificing deep consumer understanding for the sake of efficiency alone… Machines don’t experience products or form opinions like humans do.”

This hybrid approach has become a critical differentiator for Ipsos and is proving particularly effective in culturally diverse and fast-moving markets like India.

Speed, Efficiency, and Deeper Insights

Brands are now launching products in weeks rather than months, due to the increased acceleration of the product testing lifecycle as brands move from reactive to proactive innovation. FMCG companies can now predict trends and expedite development thanks to AI.

Despite digital advances, there’s a growing demand for in person product testing. The fact that 65% of Ipsos’s global product testing is still face-to-face indicates a pushback against digital-only methods. In a world increasingly shaped by algorithmic content, clients now prioritise authentic, observable consumer reactions.

The advantages of using AI into product testing are substantial from a quantitative standpoint. GenAI-enhanced concepts have a 10 percent higher trial rate, suggesting that buyers are more inclined to think about buying these goods. Additionally, approval rates for product evaluations have improved by 9–10 percent, indicating greater consumer resonance. Apart from efficacy, the use of AI has resulted in significant efficiencies. Testing cycles are now 10–50 percent quicker, and cost reductions range from 20–60 percent, allowing businesses to test more concepts within the same timeframe and budget.

Shifting KPIs: From Purchase Intent to Brand Loyalty

Another major shift is how success is being measured. Short-term purchasing intent is no longer enough for brands. They are searching for more profound markers, such as emotional resonance, long-term loyalty, and superior products.

“Clients are increasingly aware of the say-do gap,” explains Reynolds. “They want KPIs that reflect long term behaviour, not just one-time claims.”

This trend is particularly prevalent in APAC economies, such as India, where customers seek to establish long-lasting emotional and practical ties.

What’s Next? Product View AI and the Age of Agentic Systems

In the near future, Ipsos is placing a significant bet on AI's capacity to mimic human senses. The company is set to introduce Product View AI, a cutting-edge system trained in semiotics, anthropology, neurology, and ethnography. It "watches" product trials and analyzes body language, tone, surroundings, and other factors, just like a panel of experts would.

“It’s like having an ethnographer, neuroscientist, product developer and other experts in the room, analyzing you without being in the room,” says Dr. Reynolds. Through this initiative, fewer direct questions will be asked, saving time and allowing more natural consumer behaviour to emerge, yielding richer insights.

Human Truth Must Lead

Over the years, Ipsos has worked to make a strong case for placing human truth at the core of research. Reynolds focused on the concept of the “vanillization of insights,” warning of a scenario where AI systems trained on AI-generated data lead to increasingly uniform and indistinct outputs.

By trying to identify and understand the “why” behind behaviour, human elements remain irreplaceable. “HI + AI isn’t just a slogan.” says Reynolds. “It’s how we ensure our insights are not only faster but also sharper, deeper, and truly human.”

Ipsos is making sure it maintains its competitive edge as AI transforms market research. The solution is to cleverly combine speed and content rather than picking one. Real consumer understanding must always lead.