The industry has never had more ways to produce insight, and to do it faster, but that capacity isn’t the problem; rather, it’s what to do with all that data.

I spent two days at Quirk’s NYC 2026, sitting in on close to ten sessions. And when I pulled my notes together afterward, one tension kept showing up, whether the session was about AI moderation, B2B research, or sample quality. It’s that more research doesn't automatically mean better decisions. Here’s why.

The Cost of Scaling Research Without Scaling Trust

Speaker after speaker described how much capacity AI has unlocked in market research. In "How HubSpot Scaled Its Research: 100+ Interviews in Days, Not Months," they described running well over a hundred interviews in the time it used to take them to run a handful. Other examples included small customer insights teams elsewhere using AI and searchable repositories to do work that used to require a much bigger headcount.

The sheer volume of behavioral, transactional, and digital data available to researchers now dwarfs what we had access to even a couple of years ago. This is a real gain, but several sessions doubled as a warning label for exactly this trend:

• More interviews don't help if the sample behind them is biased.
• More data doesn't help if nobody trusts it.
• More findings don't help if the people receiving them can't tell what actually matters.

A few sessions pushed this conversation further upstream. The question isn't only whether you can clean bad data after collection; it's whether you created the right conditions for good data in the first place. This means the right participants, the right reasons to participate, and a sample that actually reflects the people you're trying to understand. More volume only compounds the problem if those fundamentals are weak. The goal isn't maximum research volume, but rather maximum usable value.

That's the data volume vs data value problem in a nutshell, and it's one of the clearer market research trends 2026 has surfaced. This tells me the future of this industry lies in creating more value from the insight we already have, instead of producing more just because we can.

Turning Customer Insights Into Real Influence

A few sessions took a direct swing at that value gap by asking what research is supposed to produce in the first place.

Autodesk framed it as getting the voice of the user "into the bloodstream of the organization," through workshops, videos, podcasts, and ongoing co-creation. Taken together with a few other sessions, there's a progression worth naming:

1. Data becomes a finding
2. A finding becomes an insight
3. An insight becomes a story
4. A story creates influence
5. Influence shapes a decision

The closer a customer insights team gets to that last step, actually shaping a decision, the more valuable they become.

Veterans United Home Loans made a similar point from an unexpected angle, in a session called "On with the show: What Broadway can teach us about consumer insights." You can walk into a room with twenty solid findings, but if you can't present them in a way that connects with the people you're talking to, none of it matters. It won't get used or remembered, becoming an afterthought instead of a decision driver.

While speed and scale can get you more findings faster, somebody still has to decide what those findings mean and how to say it so it actually shapes a decision. For us at Phase 5, that's the part of the job that makes everything else worth doing.

That same Veterans United Home Loans session also made a point about AI in market research, in that it can take over a lot of what researchers have traditionally done. Still, the real work, understanding people, must stay with us as humans. A session from the American Medical Association, "Insights at the edge: Building external sensing capabilities in mission-driven organizations," added a related warning. If researchers spend more time alone asking AI questions instead of talking with colleagues, the field risks losing some of the cross-functional back-and-forth that produces the best insight in the first place.

The Next Market Research Trend Is Always-On

If there's a market research trend that ties the last few years of AI hype to what's coming next, it's that research is moving away from something that starts when someone submits a request, and toward something that's always running in the background.

I heard three different versions of that idea across the conference:

• Always-on listening. HubSpot described continuously tracking how customer needs shift instead of waiting for the next scheduled study.
• Always-on relationships. Autodesk described maintaining an ongoing community of users instead of recruiting from scratch every time.
• Always-on sensing. The American Medical Association described continually monitoring signals outside the organization so insights can surface questions leadership doesn't yet know it should be asking.

None of that means always conducting studies. It means always listening, always connected, and always sensing, so that when a decision needs to be made, the organization already has a head start.

Knowing What to Do With What You Have

The one thing I'd want a client to take away from Quirk's NYC this year is that more data on its own won't get you further. Knowing what to do with it will. The advantage comes from knowing which data to trust, how to shape it into something a stakeholder will actually act on, and building the always-on habits that mean you're not starting from zero every time a question comes up.

Veterans United Home Loans made this point too, this time about squeezing the juice out of work already done. Their team talked about revisiting past studies and repackaging the same findings for different internal audiences, rather than commissioning something new every time a question comes up.

A session on small research teams reinforced the same idea from another angle, treating the insights function as an information hub rather than just a research execution function: make prior knowledge searchable, reusable, and broken into pieces people can actually find. Before asking what new research is needed, the better first question may be what the organization already knows, and whether it's actually being used.

That's a good problem for the industry to have as the emphasis is moving from how much data we can collect to how well we put it to use.

It's the same balance we try to strike at Phase 5: using AI to move faster without losing the judgment that turns data into a decision someone can actually act on.

If you're thinking about how these shifts apply to your organization, we'd welcome the conversation. Connect with Phase 5 to learn more.

Recommended Resources:

• WEBINAR - Build the Right Product Before You Build the Product: How Early-Stage VOC Drives Better Decisions
• WEBINAR Replay - Jobs-to-be-Done in the Age of AI: Finding the Next Customer Opportunity
• Size Matters in CX: What the Customer Experience Strategies Summit Revealed About the Future of CX