Turns out humans lie. Especially when they’re trying to be nice. A real-world guide to extracting useful truths without feeling like a used car salesperson.

There’s a particular optimism you have before your first proper customer interview. You imagine thoughtful answers, honest confessions, maybe even a spontaneous ‘thank you for building this, we’ve been waiting all our lives’. What you usually get instead is polite nodding, vague praise, and sentences that start with ‘I guess…’ and go absolutely nowhere.

We’re told that Customer interviews are the gold standard. Talk to users. Listen deeply. Let insight rain down upon you. The internet is full of neat templates and earnest advice that suggests if you just ask the right questions, truth will appear like a well-trained labrador.

Reality check. Customers are humans. Humans are messy, conflict-avoidant, occasionally forgetful, and strangely invested in not hurting your feelings. They lie, not maliciously, but socially. And if you don’t know how to listen for what’s real, you’ll walk away with a notebook full of nonsense and a dangerous amount of confidence.

After many years of interviewing founders, users, buyers, non-buyers, and the occasional person who definitely should not have been on the call, here are five things we wish we’d known earlier. Plus one thing that still catches us off guard, no matter how many interviews we’ve done.

Niceness is the enemy of truth

The first lie you’ll encounter isn’t ‘I’d definitely pay for this’. It’s something far more subtle. It’s politeness. Most people want to be agreeable. They don’t want to look ignorant, negative, or difficult. So they smooth the edges. They hedge. They compliment your thinking before quietly sidestepping the hard bit.

You’ll hear phrases like ‘That makes sense’, ‘I can see how that would be useful’, or the deeply dangerous ‘Yeah, I’d probably use something like that’. None of these mean what you want them to mean. They’re conversational cushions, not commitments.

Early on, we mistook these responses for validation. We left calls buzzing, convinced we were onto something big. Then nothing happened. No signups. No follow-ups. No actual behaviour to match the warm words. It took a while to realise that many failed products were built on what customers said they wanted, a pattern well documented in post-mortems of unsuccessful launches across industries.

The fix wasn’t to become aggressive or interrogative. It was to create permission for discomfort. We started explicitly saying things like ‘You won’t hurt our feelings here’ or ‘It’s actually more helpful if this doesn’t make sense to you’. Something shifts when people realise they’re allowed to be honest without social penalty.

The best interviews often feel slightly awkward. There are pauses. People think. They correct themselves. If everyone sounds confident and cheerful the whole way through, you’re probably collecting compliments, not insight.

Opinions are cheap, behaviour is gold

If there’s one mantra we now repeat like a slightly deranged monk, it’s this. People are terrible predictors of their own behaviour. Ask them what they would do, and they’ll give you an answer that reflects who they want to be, not what they actually do.

‘I’d definitely switch tools if it saved time.’
‘I’d pay for this if it existed.’
‘I care a lot about security.’

Maybe. Maybe not. The graveyard of failed products is filled with founders who built exactly what people said they wanted, only to discover that good intentions don’t survive contact with real workflows.

The breakthrough came when we stopped asking about the future and started interrogating the past. Not in a creepy way, but in a specific, grounded one. ‘Tell me about the last time this happened.’ ‘What did you do when that tool broke?’ ‘How did you solve this problem yesterday?’

Behaviour leaves footprints. Opinions float. When someone describes real actions, you get texture. Trade-offs. Frustrations. Workarounds that shouldn’t exist but do. That’s where products are born, not in hypothetical enthusiasm.

One conversation that sticks with us involved a buyer who insisted cost was their biggest concern. They spoke passionately about budgets and procurement pain. Then, casually, they mentioned paying $3,000 a year for a workaround spreadsheet because ‘it just worked’. That single aside told us more than twenty minutes of stated priorities.

Your questions are probably leading without you realising

This one stings because it’s easy to think you’re being neutral when you’re absolutely not. Especially if you’re excited about an idea. Especially if you’ve rehearsed the pitch in your head. Especially if you’re secretly hoping they’ll confirm what you already believe.

‘Wouldn’t it be helpful if…’
‘Do you think this would save time?’
‘How much do you like the idea of…’

Each of these gently nudges the other person toward agreement. Not because they’re convinced, but because disagreement feels like pushing back on you personally. Humans are weird like that.

We learned this lesson the hard way by recording and transcribing interviews, then reading our own questions back later. It was humbling. We weren’t interviewing so much as auditioning our ideas in question form.

The shift was learning to ask dumb, open questions and then shut up. ‘Can you walk me through how you do this today?’ ‘What’s annoying about that?’ ‘What happens next?’

Silence is your friend here. People rush to fill it, and in doing so, often reveal far more than they intended. The best insights usually arrive after the obvious answer, when the interviewee circles back and adds ‘Actually…’

If you feel like a therapist with a notebook, you’re probably doing it right.

You’re not the target audience (and that matters more than you think)

This sounds obvious, yet it trips people up constantly. Especially experienced operators. Especially marketers. Especially anyone who’s been in the industry a long time and thinks they ‘get it’.

Your mental model of the problem is not the same as theirs. Your language is different. Your sense of urgency is different. What feels like a glaring inefficiency to you might be background noise to them, and vice versa.

We once watched a founder interview users while subtly correcting their terminology. The users nodded along, adjusted their language, and the whole conversation quietly drifted away from how they actually thought about the problem.

That’s the danger. People will meet you where you are if you let them. And in doing so, they’ll stop showing you where they actually live.

The fix is humility. Let them define the problem space. Mirror their words back to them. If they call it ‘a mess’ instead of ‘a workflow bottleneck’, stick with ‘a mess’. That phrasing carries emotional weight. It tells you how it feels, not just how it functions.

The goal isn’t to sound smart. It’s to understand the world as it really is, not as your slide deck describes it.

One interview proves nothing

This is the part nobody likes, because it’s boring and inconvenient. One great interview does not equal truth. Neither does three. Patterns matter, not anecdotes.

We’ve all been there. One person says exactly the thing you hoped to hear. It’s tempting to treat that as a signal, especially when you’re tired, under pressure, or desperately want clarity. But outliers are seductive liars.

Real insight emerges when the same frustration shows up in different words, from different people, in different contexts. When the same workaround appears again and again. When multiple people independently complain about the same moment in the process.

This is also why note-taking discipline matters. Not verbatim transcripts, but structured summaries. What problem were they trying to solve. What triggered the pain. What did they do instead. How did they feel about it.

When you lay ten interviews side by side, themes jump out. And just as importantly, false signals fade away.

The thing that still surprises us

Even after all this, there’s one thing that still catches us off guard. People often don’t know why they do what they do. They’ll give you a reason, but it’s usually a story, not the root cause.

Dig a little, and you’ll find habits, shortcuts, internal politics, fear of change, or incentives that make no sense on the surface. A tool isn’t used because ‘it’s industry standard’. It’s used because no one wants to retrain the team. A process exists not because it’s efficient, but because it protects someone from blame.

The surprising bit isn’t that these forces exist. It’s how rarely they’re acknowledged, even by the people living with them. You can only uncover them by gently circling the issue, asking follow-ups, and being patient.

This is where customer interviews stop being a tactic and start being a craft. You’re not extracting answers. You’re uncovering context. And context is where strategy actually comes from.

Wrap-up or TL;DR

Customer interviews aren’t about asking better questions so much as listening better. People will lie to be kind, predict badly, and borrow your language if you let them. The job is to create space for honesty, anchor everything in real behaviour, and look for patterns over time. The biggest insights often arrive sideways, in asides and contradictions, not in the answers to your neat list of questions.

Do this well, and interviews stop being performative research theatre and start becoming one of the few genuinely reliable inputs into product and marketing decisions. Messy, human, and quietly invaluable.

Want to get ahead? Try doing fewer interviews, but doing them properly. You might be surprised what finally starts to make sense.