Ed Keller's recent piece in Greenbook made an argument we've been watching play out in real time for the past several years: AI is not replacing market research professionals — it's replacing the parts of their jobs that were never the point to begin with. We agree with that framing completely. But we want to add something to it, because we've been watching this shift from a vantage point that doesn't get talked about enough: the recruiting side.
EpikInsight sources, screens, and places consumers into research studies on behalf of research firms nationwide. That means we sit at the intersection of what researchers need and who can actually deliver it. And what AI is doing to how studies get fielded has changed our work significantly — in ways that are worth naming out loud.
The Studies Are Moving Faster. The Humans Still Have to Show Up.
A few years ago, a qualitative study might take weeks to field. Screeners were built manually, sent through panels, and recruiters like us spent hours on the phone validating respondents. Today, AI has compressed the front end of that process dramatically. Study designs are faster. Screener logic is more sophisticated. Researchers are asking sharper qualifying questions because they've had help thinking through what actually matters.
What hasn't changed is that someone still has to find the right person, confirm they're a real human being with genuine opinions and lived experience, and make sure they show up. That part is still ours. And if anything, the bar has gotten higher — because when researchers can field faster, they expect the humans on the other end to be higher quality.
AI Is Surfacing Better Questions. That Creates Better Participants.
Keller's point about curiosity resonates with us directly. As AI takes over the mechanical work of research — fielding, cleaning, preliminary analysis — the researchers we work with are spending more time asking harder questions. What kind of respondent actually illuminates this problem? What lived experience do we need in the room? What demographic nuance matters here versus what's just noise?
That specificity makes our job both harder and more interesting. We're not just matching zip codes and age ranges anymore. We're being asked to find people whose actual lives reflect the research question. That takes human judgment. It takes real conversation. It takes a recruiter who understands what a focus group is actually trying to accomplish — not just one who can work a screener.
When the research itself gets smarter, cheap panel fill becomes a liability. Researchers notice when the respondents don't match the brief.
The Commodity End Is Gone. Good.
If AI has done anything for the participant recruitment space, it's exposed how thin the value proposition was for firms that were just running bodies through panels without any real vetting. When the research itself gets smarter, cheap panel fill becomes a liability. Researchers notice when the respondents don't match the brief. They notice when someone clearly qualifies on paper but has nothing to say in the room.
That's the part of this industry we've never been comfortable with. EpikInsight was built on a different premise: that the respondent experience matters, that honest screening beats volume, and that a well-placed participant is worth more than ten warm bodies who technically clear the filter.
AI is making that distinction harder to ignore. We're fine with that.
What This Means Going Forward
Keller ends his piece with a call to action: invest in curiosity and creativity, because those are the skills that AI can't replicate. We'd extend that to the recruitment side of the industry. The firms that will matter going forward are the ones that understand what they're actually placing — not just who qualifies on a checklist, but who has something genuine to contribute to the research.
That takes industry knowledge. It takes the kind of judgment that comes from years of conversations with real people about what they actually think and feel and buy and use. It takes, frankly, the same curiosity that Keller is asking of insights professionals.
We've watched AI take over the parts of this work that should have been automated a long time ago. What's left is the part that was always the point.