The $20,000 Focus Group Is Dead

In 2019, I watched a Fortune 500 brand write a $22,000 check for a focus group. Six weeks later, they got a PDF. Forty-three pages of carefully formatted insights, color-coded charts, and bullet points summarizing what twelve people in a room in Cincinnati thought about a new line of cleaning products.

The product launched anyway. It underperformed. The insights were "directionally correct" — the consultant's words, not mine.

I've thought about that moment a lot over the past few years. Not because of the waste — though it was staggering — but because of the certainty with which everyone in that room treated those forty-three pages. As if the opinion of twelve carefully recruited suburbanites, speaking into a one-way mirror while eating free sandwiches, was the clearest possible signal of what a market of millions would do.

It wasn't. And we all knew it. We just didn't have a better option.

Now we do.

The Real Cost of Traditional Focus Groups

Let me break down where that $20,000 actually goes, because the sticker shock deserves an itemized receipt:

• Recruiting fees: $2,000–$5,000. Finding twelve people who fit your screener criteria, live within driving distance of a facility, and can show up on a Tuesday afternoon at 2pm isn't easy. Agencies charge accordingly.
• Participant incentives: $75–$150 per person, plus travel reimbursement. Call it $1,800 on a good day, closer to $3,000 if your screener is tight.
• Facility rental: $1,000–$3,000 for a half-day in a proper focus group suite — the kind with the one-way mirror, the comfortable chairs, the streaming setup for remote observers.
• Moderator fees: $3,000–$6,000 for a skilled qualitative researcher who can keep a group on track, probe for depth, and not lead the witness.
• Analysis and reporting: $3,000–$6,000 more to turn eight hours of recordings and transcripts into something a VP will actually read.
• Time: And then there's the invisible cost nobody puts on the invoice. From brief to final report, you're typically looking at four to eight weeks. In a fast-moving market, that's an eternity.

Add it up and you're looking at $10,000 on the low end for a single-city, single-session study. Multi-city? Multiply accordingly.

What You Actually Get

Here's the part that kept me up at night even before AI made it obsolete: traditional focus groups are structurally compromised from the moment you start recruiting.

And a sample of n=12 carries real statistical risk, especially when you layer in the biases that naturally emerge in group settings:

• Social desirability bias: People say what they think sounds good, not what they actually feel. Nobody wants to be the person who admits they wouldn't pay $45 for a moisturizer when the woman next to them just said she would.
• Groupthink: Strong personalities dominate. Mild-mannered participants fall in line. You end up with a consensus that reflects the loudest person in the room, not the actual distribution of opinion.
• Recall bias: Participants are asked to remember how they felt about something they've never encountered before, which means they're constructing opinions in real time rather than reporting them.
• Moderator effect: Even the best moderators introduce subtle cues. Tone, follow-up questions, body language — it all shapes what people say.

None of this makes traditional focus groups useless. But it does mean that a $20,000 study might be giving you a funhouse-mirror version of reality — expensive, slow, and potentially misleading.

What AI Changes

AI-powered synthetic focus groups like PROVA don't eliminate research bias entirely — nothing does. But they change the economics and the speed so dramatically that they make iteration possible in a way that traditional research never did.

Here's what changes:

• Sample size becomes unlimited. Run ten personas or a hundred. The cost difference is negligible.
• Speed becomes irrelevant. A PROVA session takes five minutes to run and another two to read the output. You can run three variations before lunch.
• Iteration becomes affordable. Want to test a revised concept after reading the first round of feedback? Run it again. Immediately. For the same price.
• Demographic targeting becomes precise. Build a panel of exactly the consumers you want — age, gender, location, income, buying behavior, brand affinity — without paying a recruiter $3,000 to find them.

The Math Is Brutal

Let me put it side by side so the contrast is unmistakable:

Dimension	Traditional Focus Group	PROVA AI Focus Group
Cost per study	$10,000–$22,000	$99/month (unlimited)
Time to results	4–8 weeks	5 minutes
Sample size	8–12 participants	Up to 15 per session (unlimited sessions)
Iterations	1 (maybe 2 if budget allows)	Unlimited
Geographic flexibility	One city per study	Any demographic profile
Social desirability bias	High (group setting)	Low (AI personas respond independently)
Recruiter fees	$2,000–$5,000	$0
Analysis time	1–2 weeks after session	Instant, included in session

Is It As Good?

This is the honest question, and it deserves an honest answer.

No. AI synthetic focus groups are not a replacement for longitudinal ethnographic research, for deep-dive in-context observation, or for the kind of qualitative insight that only comes from watching a real human being struggle to open your packaging for the first time. For those use cases, the traditional methods still have value.

But here's the thing: for 95% of the decisions most teams are actually making — which tagline to test, which price point to run, whether a concept resonates before you spend $200,000 on production — PROVA delivers directional clarity faster and cheaper than any alternative.

And directional clarity, delivered in five minutes at $99 a month, beats forty-three pages of carefully formatted uncertainty at $22,000 every single time.

The $20,000 focus group isn't dead because AI is perfect. It's dead because the trade-off no longer makes sense.

If you're still writing those checks, I'd gently suggest you run one PROVA session first. Just to see. The worst case is you come out more confident in the decision you were already going to make.

The best case is you save $22,000 and six weeks.