I recently sat through eight conference talks that left me with that rare, uncomfortable feeling: the industry is finally saying out loud what many of us have known for years.
The thread running through all of it? Most organizations are still measuring the past and calling it insight. Tracking studies that report on what happened. Surveys that capture what people say, not what they do.
As someone who works with insight and CX teams every day, I find this both vindicating and urgent. Because the tools to do better aren't theoretical anymore. They're deployed, they're working, and the brands using them are moving faster than the ones still defending legacy methodologies.
Here's what stood out from the conference, and what it means for market researchers who want to lead rather than lag.
1. AI isn't a feature, it's becoming research infrastructure
Multiple sessions showed AI being used to run qualitative interviews, generate digital consumer twins, and power always-on tracking. Not as an experiment, but as a core methodology.
What this signals for insight managers is clear: the always-on model that communities were built for is now the default expectation.
Communities truly give brands an edge. They aren't waiting weeks between research cycles. On the other hand, they are maintaining an engaged, opted-in panel of real customers. Already profiled, already consented, already warm & ready to respond the moment a business question emerges. And crucially, the ability to profile your customers over time means that understanding of your members & how they evolve deepens over time. Every interaction adds a layer: attitudes, behaviours, life stage, category usage. You don't have to ask everything upfront. You build a richer picture with every touchpoint, without survey fatigue eroding your response rates.
What to prioritize: Move away from thinking about research as projects. Start thinking about research as a program. With your insight community as the backbone of your research program that makes continuous, contextual insight possible. AI complements your research, helps you scale your insights but your communities help you with authentic feedback to keep a tap on the pulse of your customers.
2. Mental availability is reshaping what brand tracking is actually for
One of the most compelling sessions I attended, focused on how brands are moving beyond awareness measurement towards building associative memory structures — staying top of mind in the specific moments that trigger purchase decisions. One European footwear brand discovered it had virtually no presence in the gym occasion as a buying trigger, despite assuming otherwise. That gap only emerged because someone had longitudinal data to see what was missing.
This is exactly where insight communities create structural advantage. Unlike ad-hoc studies that give you a snapshot, your community tracks how associations shift over time, across segments, in response to real-world events. And when you need to go deeper on the why, you can leverage qualitative tools & let your members show you their experience in context, not just describe it in a survey. Screen recordings, task completion, verbal walkthroughs: the kind of evidence that makes brand and product teams sit up.
Tip for market researchers: Don't just track top-of-mind awareness. Map your questions to buying occasions and emotional triggers. Ask your community members to walk you through their last purchase decision in a category. The signals you find will be worth more than a thousand prompted recall scores.
3. The foresight turn is real & backward-looking research won't survive it
Unilever identified the opportunity for a blue light protection cream not through category surveys, but through behavioural observation in niche skincare communities. The places where early adopters were already problem-aware, long before mass-market awareness caught up. O2 made a similar argument: confirmation bias and the say-do gap aren't occasional errors, they're structural features of how most insight programmes are designed.
Your community is already a foresight asset. Most teams just aren't mining it that way. Your most engaged members are often early adopters, category enthusiasts, and vocal opinion formers. Open-ended discussion boards, diary studies, and ethnographic tasks let you listen for emerging signals rather than just measuring against pre-set hypotheses.
For teams that want to go further with advanced methodologies, tools such as MaxDiff and Conjoint analysis help you to move from "what do people say they prefer?" to "what do people actually trade off when it comes to a real decision?" These aren't niche techniques anymore. They're the methodologies that separate insight that drives product and portfolio strategy from insight that produces a nice-looking chart.
What to prioritise: Schedule exploratory, unstructured community touchpoints alongside your structured surveys. Ask open questions. Use AI to uncover prompted follow-ups to go deeper. Look for the things people bring up that you didn't ask about. That's where foresight lives.
4. The bottleneck isn't knowledge. It's distribution
Two sessions made the point clearly: organizations aren't short on research. They're short on stakeholders who act on it. The insight team's real challenge is getting findings into decisions at the speed those decisions are being made.
This is where organisations need a solution that powers enterprise grade controls and seamlessly allows you to scale access. Rather than every insight request going through a single research team, you want a solution that allows you to configure who can access which data, run which methodologies, and view different profiles/ segments with appropriate governance guardrails in place. The vision would be to enable marketing to self-serve a quick poll. Product to run a concept test. In this way, the insight team retains oversight and quality control, while the organisation moves faster.
And when it comes to bringing everything together, you don’t want to export data from one tool, stitch it with SPSS output, and cross-reference a qual platform that lives somewhere else entirely. Quantitative results, qualitative themes, conjoint outputs, UX findings. You need AI to help you analyse everything in one place, by teams who can actually see the full picture.
Tip for market researchers: Stop measuring your success by the quality of your decks. Start measuring it by how many decisions were made differently because of what you told someone. The organisations winning here have moved insight from a service function to a strategic nerve centre.
5. Global reach and sample quality are becoming competitive advantages
One of the sessions highlighted how passive and behavioural data is raising the bar on what "valid" research looks like. Opt-in panels were challenged. Self-selecting respondents were challenged. The implicit assumption that survey data reflects real behaviour was challenged loudly.
This is fair. But the answer isn't to abandon communities — it's to be honest about what they're for, and to pair them with access to representative sample where needed.
While your communities power a single source of truth, for global brands and research teams working across markets, access to international sample means you can field studies across geographies without spinning up separate panels in each market or relying on a patchwork of regional vendors. Whether you're validating a concept in Southeast Asia, tracking brand health across Europe, or running a pricing study in the US and comparing it to Latin America, the infrastructure is in the community and should be seamlessly topped up with representative sample where required.
What to prioritise: Think about your insight programme as layered. Your community gives you depth, longitudinal tracking, and the qualitative richness that makes findings actionable. Sample access gives you representativeness and geographic breadth. This combination is more powerful than either alone.
The through-line and what it demands of us
Every one of these trends points to the same shift: research programs built for coverage and comfort are being replaced by ones built for signal quality and strategic relevance.
That's not a threat to insight communities, it's the strongest argument for them. But only if we use them properly.
The brands doing this well aren't running their communities as a cheaper, faster survey panel. They're building longitudinal relationships with the people who matter most. They're using progressive profiling to deepen their understanding without asking the same questions twice. They're running MaxDiff and conjoint studies when decisions warrant rigour, and unmoderated UX tasks when they need to see behaviour, not just opinions. They're democratising access so the whole organisation moves faster and they're analysing everything in one place so nothing falls through the cracks.
The question for every market researcher reading this: Is your insight program built around what's easy to measure, or around what's actually worth knowing?
