The Problem with Customer Feedback

There's a problem with customer feedback that has nothing to do with bad research.

It's quite simply this: Your customers chose you.

Obvious, maybe. But we rarely account for what that choice already filtered out.

Think about who tends to sit closest to a product team.

Current customers. Power users. Advisory-board members. Champions inside key accounts. People willing to join a beta. Customers who take your calls.

These are often exactly the people worth talking to. They know the product. They understand the category. They can name the friction, the value, where competitors are stronger, and what would make the experience better.

But they share something else.

At some point, the existing proposition already cleared their threshold. They bought. They adopted. They stayed. They got proficient enough to build processes, careers, or internal credibility around your product.

That history gives their evidence a point of view.

And the point of view matters most when the question is bigger than, "How do we improve what we already have?"

Say you're considering a real shift. A new workflow. A new tier. An AI capability. A move into an adjacent use case. A change to the basic value proposition.

You take the concept to twenty current customers. They get it quickly. Several are enthusiastic. A few say they'd use it tomorrow. The advisory board sees the potential. Your largest customer says it fits where their organization is headed.

That's useful evidence. It's also less of a test than it might appear.

The people responding well have already shown they're compatible with how your company sees the problem. They accepted your category assumptions, your tradeoffs, your architecture. Or enough of them to become customers in the first place.

So, when they react to the next idea, you're testing it among a population that already passed through a filter. The bet has, in some form, already worked on them.

That's why I create deliberate distance from the installed base when a product premise carries real consequence.

I want people who use competitors and have good reasons for it. Former customers who can explain why the relationship stopped working. Experienced prospects who looked hard at the category and passed. People who understand the problem well enough to recognize the idea's logic, but haven't already committed to your version of the answer.

Within that group, I'm especially interested in Catalytic Customers.

They aren't valuable because they're contrarians. They're valuable because their disagreement is informed. They've seen enough of the category to understand the tradeoffs, compared enough approaches, and have the access and role to explain why something would or wouldn't work. They tend to be constructively critical — not reflexively positive, not reflexively negative.

They can look at a promising idea and say:

I understand what you're solving for, but this problem isn't big enough to make me change.

Or:

That works if the buyer and user are the same person. In our organization they aren't, and you haven't accounted for that.

Or:

I'd believe this if you could do X. Without it, the switching cost outweighs the benefit.

Those aren't objections to overcome. They're information — often information you'll never get from the people most inclined to support the idea.

This is where research creates false confidence. You can run the interviews well, recruit real customers, ask neutral questions, get consistent reactions, triangulate across studies and still end up with evidence that never met resistance.

The problem isn't a function of rigor. It's a matter of range.

Who had a credible reason to see the premise differently? Who knew enough to challenge the assumptions underneath it, or to say the existing way was good enough? Who had already considered this kind of change and turned it down?

If those voices are missing, agreement starts to look more independent than it is. Five studies of the same customer population aren't five tests of the premise. Sometimes they're the same lens, five times.

That has consequences because the market isn't made up entirely of people who already chose you. Growth depends on persuading competitor users, light users, former customers, skeptical prospects, new buying centers — people whose current behavior says they don't yet believe what you believe.

They're not an edge case. They're often where the real decision lives.

None of this means walking away from current customers. Installed-base research is essential. Your customers tell you things outsiders can't.

The mistake is asking one population to answer every question.

Current customers are exceptionally good at telling you how to improve what they already have. They're not always the population I'd trust most to tell me whether the next bet deserves to exist.

For that, I want a harder test: people who know the category, have the experience to judge it credibly, and no particular reason to reassure me.

The enthusiast can tell you why an idea is appealing. A Catalytic Customer can tell you whether it survives contact with someone still free to say no.

That's the evidence I want before the market gets its turn.