The Product Bet Is Only as Strong as the Resistance It Has Survived

Product teams rarely have too little customer evidence. Usually it's the opposite. There are interview transcripts, survey results, win-loss findings, usage data, support tickets, advisory board notes, customer calls, and a few internal people who just know the market cold. By the time a real roadmap decision reaches senior leadership, the case behind it can be thick.

And some of those decisions are still fragile as hell.

They get reopened. Sales walks in with a customer objection. Someone finds a new data point. The CEO asks if the team's actually sure. Someone proposes another round of research, and a decision that felt settled a month ago is suddenly back on the table.

That should tell you something. The issue probably isn't that the team didn't do the research. It's that the research never really tested whether the premise underneath it could hold.

Agreement feels like certainty, but it isn't

Research has a way of becoming more reassuring than it should be. A team starts with a reasonable hypothesis about a customer problem. They talk to customers who recognize it. Those customers describe the pain, walk through their workaround, and light up a little when they see a proposed fix.

It starts to look like a slam dunk.

The tricky part is that none of that evidence has to be wrong. The customer really does deal with the problem. The workaround really is annoying. The proposed fix really would help. What's still an open question is whether any of that is worth building for.

The customer who says "I'd definitely use that" is easy to listen to. The customer who says "the old way's fine" is easy to wave off. Maybe they're not the target user. Maybe they don't get the vision. Maybe they just hate change.

Sometimes that's true. And sometimes that customer just put their finger on the weakest part of your case. They might know the current workaround is good enough. That the pain doesn't come up that often. That switching costs more than it's worth. Or that this problem is way down the list of things actually competing for their attention.

A customer who pushes back isn't automatically right. But they're often a lot more useful than the fifth person who shows up to agree with you.

A real problem isn't the same as a problem worth solving

This is where a lot of product research starts to lose the plot. Proving a problem exists is not that hard. Customers can describe frustration on command. They'll remember a recent example. They'll gripe about the current setup, and if you show most of them a better version, they'll say yes, that's better.

That's evidence a problem exists. It is not evidence you've found a strong opportunity.

You still need to know how often it happens, how much it actually costs the customer, how they're coping with it right now, what else is ahead of it in line, and what they'd actually be willing to change to fix it.

There's a big gap between "yeah, that bugs me" and "I need this fixed." And the money lives entirely on the far side of that gap. Plenty of products make something a little better. Far fewer products get customers to change behavior, switch tools, pay more, learn a new workflow, or actually pick one option over another and stick with it.

That's what makes a positive reaction to a concept so easy to misread. Customers can tell an idea is better than what they have now without thinking it's important enough to act on.

Why some decisions just won't stay decided

This is also why certain roadmap calls keep coming back around. If the original research proved customers liked the idea but never pushed on the edges of that argument, different people in the room end up reading it differently.

Product sees unmet need. Sales sees customers with other priorities. Finance isn't sold on willingness to pay. Leadership sees an opportunity but can't tell if it's worth the engineering time.

Nobody's necessarily wrong. Everyone's looking at real evidence. The decision stays shaky because the evidence never forced the underlying assumptions to actually fight each other.

That's a different problem than getting stakeholders aligned. You can get people aligned around a half-built argument for a while — that part's easy. The harder job is building an argument that can take a real hit and still stand. And one of the best ways to do that is to go looking for the customer who's going to throw the punch.

What a tougher customer actually gets you

This is basically the whole idea behind what I call Catalytic Customers.

Catalytic Customers are experienced, engaged, well-informed people in the category who are also willing to be a pain about it. They're not necessarily your biggest accounts. They're not power users or early adopters or the people who post nice things about you online. And they're definitely not picked because you think they'll like the idea.

What makes them worth the time is the quality of their judgment. They've usually got enough scar tissue to compare your product against real alternatives. They understand how the category actually works, not just how your team talks about it. They can tell the difference between something that looks like an improvement and something that actually changes anything.

And they'll tell you no. They might tell you a feature solves a problem they barely notice. They might tell you your big differentiator is already becoming the price of entry. They might say the workflow you're trying to optimize shouldn't exist in the first place. Or they'll tell you your instinct is right, but you're fixing the least important five percent of it.

None of that feels great in the room. It's usually exactly what the decision needed.

Research should stress-test the case, not just support it

There's a habit of treating customer research as evidence collection — go find enough proof and move on. I think that's setting the bar too low.

For the decisions that actually matter, research should also try to break the argument. It should tell you not just why the idea might work, but why it might fail. It should go find the customers most likely to reject the whole premise and actually understand their reasoning. It should separate enthusiasm from importance. And it should be honest about which parts of the case are still assumptions dressed up as findings.

Good research occasionally makes a team less confident, not more. That sounds bad, especially when everyone wants to move fast. But finding out a decision is weaker than it looked before you've spent the engineering budget is a lot cheaper than finding out after launch.

The goal was never certainty. Product decisions almost never come with that option. The goal is a bet you can actually defend.

The best question to ask before a decision locks

Before an important roadmap call gets finalized, I ask one question: who's most likely to disagree with this? Then I make sure that person actually shows up in the customer evidence.

Not because they get a vote. Not because the skeptics should be running the roadmap. Just because an argument that's never met real resistance hasn't actually been tested yet.

The strongest product decisions I've seen are almost never backed by the most glowing customer quotes. They're the ones where the team already knows the limits of the opportunity, already knows exactly what customers will push back on, and decided the bet is still worth making anyway.

That's a different kind of confidence. And it's the kind that's still standing when someone challenges it three meetings from now.