AI in Market Research
Unsurprisingly, AI is having a huge impact. Some of the most exciting applications I encountered included:
- “Qual at Scale” Tools Qualitative research has traditionally been slow and time-intensive. AI tools are now automating key parts of the process—from writing discussion guides to interviewing participants to summarizing results. Two key approaches stood out:
- Conversational AI: These chatbots probe deeply into topics and can conduct detailed interviews. My personal experience as a respondent, however, found them to be a bit exhausting. Features like voice/video input and better incentives could improve engagement.
- Quant/Qual Hybrids: These combine structured survey questions with open-ended qualitative feedback. It’s a great solution when you need quantitative data with some added qualitative insights.
At Peel, we’ve been experimenting with AI in our qualitative research—primarily for analysis, but also for planning and setup. While it doesn’t replace an experienced researcher’s expertise, AI speeds up the process significantly. One delighting benefit that stood out to me: loading transcripts into an LLM creates a searchable database of interviews, making it easy to pull key quotes or verify insights without exhaustive searching.
- Bringing Personas to Life AI tools are now helping enhance segmentations and personas with richer details like lifestyle traits, attitudes, and even realistic imagery. In one example, a firm even created a video dialogue between two fictional characters based on personas. It was very cool but spooky.
- Brainstorming AI excels as a brainstorming partner—especially after a round of qualitative research, such as to develop new product opportunities, messaging platforms, claims, product names, etc. You can even have AI brainstorm ideas within different consumer segments or personas.
- Synthetic Data Synthetic data simulates additional responses based on existing data. This approach is helpful when you need insights from hard-to-reach groups or want to explore new questions using existing study data. While synthetic data offers solid accuracy (~85%), there’s ongoing debate about its role in research. Personally, I believe synthetic data is an exciting supplement—not a replacement for real sampling—and can extend research capabilities when budgets are tight or precision isn’t critical.
Brand Health Research
A new approach inspired by the “How Brands Grow” book series is gaining traction. This method moves away from traditional marketing funnels and instead focuses on building memory structures that connect brands to specific consumption occasions. We’ve started using this approach at Peel, and it’s proving highly effective. I’ll share more details in a future post, but reach out if you’d like to discuss.
Data Quality
Survey fraud remains a concern, but with the right strategies in place, it’s manageable. While panel providers are improving their defenses, new third-party tools are also emerging to screen participants. These tools help detect respondents using VPNs, auto-translating surveys, copy-pasting answers, or attempting to qualify through “yes-saying,” just to name a few. At Peel, we’ve found these services invaluable—but they work best as part of a broader quality control strategy.
I’m sure I missed some important trends, so feel free to reach out to me and tell me your thoughts! What new research methods are you most excited about?
