How AI is reshaping market research

Why AI is no longer optional

Consumers today have short attention spans and high expectations. They crave novelty and demand speed. To keep up, businesses need more than traditional research—they need AI. Take Beck’s AI-created beer in March 2023 (https://www.youtube.com/watch?v=DLxso2mBeuQ), for instance: every aspect, from recipe to marketing, was machine-generated. Though controversial, it sparked critical conversations about the future of AI in creation—and helped us see the disruptive potential when brands push boundaries.

Blending human intuition with AI intelligence

Historically, innovation followed a linear, human-led process. But AI enables a hybrid model where machine intelligence and human expertise collaborate across the funnel: from early discovery to final design and validation.

1. Discovery: Trendspotting and foresight

AI-powered platforms like AI Palette and Black Swan Data scan billions of data points to reveal emerging trends and shifting behaviours. Combined with synthetic data—AI-generated data simulating real-world responses—researchers can test early hypotheses at speed and scale, without privacy constraints.

Even more powerful are synthetic humans: simulated personas that behave like real people, built on vast behavioural and psychographic data. Using multi-agent systems — combining AI with input from psychologists and behavioral experts — these personas take part in virtual interviews to reveal deeper insights into people’s motivations, attitudes, and lifestyles. And thanks to Retrieval-Augmented Generation (RAG), these personas can be enriched with your own brand data—making them incredibly brand-smart. But here’s the catch: while synthetic users are fast and scalable, they lack true human empathy. They won’t notice a smirk or shift in tone. That’s why real conversations remain pivotal — they drive true understanding.

2. Definition: from category insights to brand archetypes

In a project for Lamb Weston, we used AI to simulate seven snacking personas in France—each informed by psychographics, behaviours, and real-world context. Our qualitative researchers conversed with them as they would in qualitative interviews, surfacing rich insights into snacking occasions, attitudes, and emotional drivers.

Synthetic humans also help us decode brand perception. In another case, we personified three Belgian fast-food brands through relatable characters. Each AI-generated persona reflected distinct brand personalities—helping us understand how Gen Z perceives and connects with different players in the market.

3. Design: visualising concepts and crafting packaging

When it’s time to bring ideas to life, AI tools like Midjourney or ChatGPT E help generate concept visuals in seconds. But having a good-looking pack isn’t enough. You need proof.

That’s where predictive eye-tracking comes in. Simulating where people look within the first 3–5 seconds, this tool shows what captures attention—and what doesn’t. With predictive heatmaps, we refine layouts to ensure real triggers will stand out. As often small design tweaks made big differences. This is also valid for shelf stand-out. But also here the human factor cannot be underestimated. Yes, these type of tools can identify gaze patterns based on contrast, shape etc. But they cannot predict what is already installed in consumers/shoppers visual memory regarding a brand or category… So, also here the combination is of vital importance.

4. Validation: testing ideas with empathy and scale

When you have found your distinctive concept, it is recommended to test it at scale via quantitative research. For this part we use conversational AI bots in quant surveys to ask smart follow-ups—adding qualitative depth without extra fieldwork. And for hard-to-reach audiences (e.g. high-income investors), we use augmented sample blending and boosting—techniques that fill data gaps or simulate entire populations based on real survey insights.  

Finally, video-based research captures emotion in a way surveys can’t. Tools like our Try & Tell platform let consumers submit product reviews via video, at scale. These are auto-transcribed, keyword-tagged, and—most impressively—made searchable and interactive via AI chat.  

5. Ready to launch—with confidence!

With AI, your innovation doesn’t just reach the market—it arrives smarter and more relevant. But success doesn’t end at validation. You’ll still need a strong business case, and by following these steps, you’ll build the foundation to gain faster stakeholder buy-in.  

To close, below some golden rules for using AI in research :

• Avoid the hype: If a tool doesn’t solve a real problem, skip it.

• Know your data: Bad input = bad output.

• Stay transparent: Understand how your tools work.

• Respect ethics: Especially with voice and video data.

• Stay human: Use AI to uncover the what, but people to understand the why.

In conclusion: AI isn’t replacing us—it’s empowering us

Used wisely, AI doesn’t diminish the researcher’s role—it supercharges it. By blending human empathy with machine intelligence, we can uncover sharper insights, spot trends sooner, and design innovations with real-world impact. Turning your innovation process into a true competitive edge. The future is hybrid. The tools are here. The superpower is yours. Are you ready to unlock it?

Send in one specific research question — for example, about a target group, brand, campaign, or product — before July 15, 2025.
The first 10 submissions will receive one free analysis based on their question.
Note: One question per organization.