AI Is Sorting Insights Teams Into Indispensable and Optional. 7 Ways to Make the Cut.
Author: Adrien Weinert, Head of AI Research, Firefish
Every insights professional I talk to is asking a version of the same question: what’s my job now?
AI in market research is changing where insights teams create value, and the industry feels it. Wherever I go at the moment, from conference stages to client meetings to colleague catch-ups, the concern is real. Roles are shifting fast, and people want to know where their value sits.
I’ve been working through this question for a while now. I’ve spoken at MRS and IIEX Europe with Unilever’s Anne-Claire Pierron about the false confidence AI can create in research, and more recently I’ve written about the messy middle, the myriad of seemingly smaller decisions that move ahead without consumer input.
Here are seven ways to make sure your insights team lands on the indispensable side.
1. AI Agrees Too Easily. Build in the Argument.
AI is a people-pleaser. How we build it into research matters as much as what it’s technically capable of. Used well, it complements, challenges, and expands human thinking.
Left alone, it tends to agree. It’s trained to be helpful, so it often validates whatever thinking it’s given. Treat it as an authority, and you inherit its blind spots.
We’ve seen where this goes wrong. Ask AI to generate conclusions, and it fills the gaps with something plausible. Ask it to summarise hundreds of responses and the outlier comment, often the most useful one, gets averaged away. The output looks finished, so nobody checks the logic. And that’s how easily false confidence gets into a business.
The solution is to build challenge in at every stage. Researchers design questions that provoke a genuine reaction. AI runs the conversation and the first pass of analysis. Experienced researchers then bring the judgement, catching what the summary missed, checking the logic before anyone trusts it and turning findings into a clear direction on what to do next. The result is insight that teams can act on with confidence, at the speed the decision needs.
That’s how AI can bring the consumer voice into the messy middle and still hold the bar high. AI is meant to amplify the work, while humans stay the author. It’s the approach behind QualifyAI, recognised by ESOMAR for Excellence in Global Research in 2025 and nominated for two MRS awards this year.
2. AI in Market Research Made Work Faster. Now It Has to Get Better.
Everyone needs everything for yesterday, so faster was the easy win. AI’s clearest impact in market research so far has been making work faster (and cheaper). In the process, it’s also made much of it flatter: the same summaries, the same safe conclusions, the same polished decks. The risk is that efficiency becomes the goal.
Speed earns its place when it brings insight to decisions that usually move ahead on instinct. That’s the messy middle: the smaller, faster calls that shape what reaches the market. They don’t wait for a full research programme, and AI-enabled research should now be used to bring the consumer voice into them in days.
The bigger prize is using AI to improve the thinking. Scenario planning, war gaming, and other consultancy-style applications stretch what insights teams can offer.
3. Teams Are Undercharging for Their Impact
Insights teams are brilliant at creating value and terrible at claiming it. This needs to change.
The strongest teams build in impact measurement from day one. They revisit older projects once the longer-term effects show, and link insight work to business outcomes and revenue in the language the boardroom already uses.
This matters more every year as AI takes on more of the core market research tasks. Insights professionals who can define, name, and price the value they provide are best placed to amplify and repeat it.
4. Researcher Out, Insight Architect In
The best researchers of the next few years will hand AI the legwork and keep the judgement. As AI takes on more of the doing, the value of insights teams moves from conducting research to designing the thinking behind it. The role moves toward consultant and strategic advisor.
An insight architect starts with the business question. They decide what evidence is needed, spot when the answer may already exist and connect different sources into a clear direction.
It all rests on knowing the data like the back of your hand. That familiarity is what lets an architect judge what AI hands back, and it’s what makes their advice worth acting on.
5. Influence Lives Outside the Deck
The deck is often the least persuasive thing in a stakeholder meeting. AI makes written output quick and abundant, so the written word carries less weight than it used to. The value comes from relationships, trust, and ensuring insight drives change.
Much of that happens around the research itself. Being in the room, the conversations before and after a debrief, and trust built over time often shape decisions more than any report.
That puts pressure on the presentation-style debrief, especially in remote teams. The standout skill is making findings memorable, sparking discussion, and “agitating” the organisation into action.
6. More Answers, Fewer Decisions
Everyone has answers now. And that’s the problem. AI is putting insight into more hands across the business, so more people can find answers themselves. That’s good news for organisations, with one catch Anne-Claire put perfectly at IIEX: having access to a skill and being skilled are two completely different things.
When everyone has their own sources, and their own AI tools, information multiplies. Teams end up with more answers than they can use, and decisions stall under the weight of it.
That’s where insights teams truly earn their place. Their value sits in coherence, cutting through the overload, adding context and judgement, and giving the business a clear basis for a decision.
7. Make Human Expertise Impossible to Miss
AI work and human work can look identical on the page, but only one of them knows when something isn’t right.
The risk that keeps surfacing is false confidence, when work looks polished enough that nobody thinks to question it. AI produces expert-looking work with no expert experience required, rarely flags its own limits, and can flatten nuance, miss the outlier or misread tone while still sounding entirely convincing. That makes human expertise more valuable, and easier to overlook.
Insights teams need to make the human check non-negotiable and visible. Show where experienced researchers shaped the question, challenged and elevated the output and caught what AI missed. Show the years of research literacy that let someone sense when something feels off. That’s what keeps confidence from outpacing competence, and it’s what stakeholders are really trusting.
Where Consumer Insights Professionals Belong
Taken together, these seven points describe one shift. Insights professionals must move more intentionally from producing research to guiding decisions: proving impact, framing better questions, orchestrating evidence, challenging thinking and creating meaning.
AI can accelerate the work in market research. But decisions still need people who can judge what the evidence means, and who have the relationships to make the business act on it. AI’s speed, paired with that human judgement, is what we call strategic intelligence: understanding that enables confident action. And that’s where insights professionals belong; in the room where decisions get made, as the judgement the business trusts. And remember, AI is helping us make that judgement more valuable every year.
Figuring out where AI’s speed ends and your team’s judgement comes in? We’re happy to think it through with you. Get in touch.