ESSAY
April 23rd, 2025
Stop Listening To Your Customers—Watch Them Instead
Stop Listening To Your Customers—Watch Them Instead
ELLIOT CRANE
ELLIOT CRANE
The product development world has turned "talk to your customers" into something of a sacred ritual. Founders dutifully show up at coffee shops and Zoom calls, notebooks in hand, eager to collect wisdom that will guide their next build. But here's the uncomfortable truth we've witnessed across countless startups: what people say in these interviews often has shockingly little connection to what they'll actually do when no one's watching. The problem isn't that customers are lying – it's that humans are notoriously bad storytellers when it comes to explaining our own behaviors and motivations. We've seen founders walk away from these conversations brimming with confidence, only to build products that solve problems nobody actually tries to fix in real life.
The product development world has turned "talk to your customers" into something of a sacred ritual. Founders dutifully show up at coffee shops and Zoom calls, notebooks in hand, eager to collect wisdom that will guide their next build. But here's the uncomfortable truth we've witnessed across countless startups: what people say in these interviews often has shockingly little connection to what they'll actually do when no one's watching. The problem isn't that customers are lying – it's that humans are notoriously bad storytellers when it comes to explaining our own behaviors and motivations. We've seen founders walk away from these conversations brimming with confidence, only to build products that solve problems nobody actually tries to fix in real life.
When you sit someone down and ask directly about their needs, they'll spin up a perfectly reasonable-sounding story that casts them as thoughtful decision-makers with clear priorities. They'll enthusiastically endorse features they'll never touch, describe workflows that bear zero resemblance to their daily habits, and make predictions about their future behavior that directly contradict what they've done for years. It's not malicious – we humans simply lack access to the messy, subconscious forces that actually drive our decisions. And the validation founders think they've achieved? Often just a mirage that feels good but offers little predictive power about market success. Things get worse when founders unknowingly structure interviews in ways that guide responses toward foregone conclusions. Ask someone about pain points in their category management process, and suddenly they'll discover frustrations they never knew they had. Inquire about automated solutions, and watch as enthusiasm materializes even from people who routinely ignore or disable automation in their current tools. Each question plants subtle suggestions about what should matter, unconsciously shaping responses that feel genuine but actually just mirror the conversation's structure. Most dangerous is how these sessions create a false sense of forward momentum – founders leave convinced they've done proper validation work and have permission to forge ahead with their existing vision, building for what customers say rather than what they actually do.
When you sit someone down and ask directly about their needs, they'll spin up a perfectly reasonable-sounding story that casts them as thoughtful decision-makers with clear priorities. They'll enthusiastically endorse features they'll never touch, describe workflows that bear zero resemblance to their daily habits, and make predictions about their future behavior that directly contradict what they've done for years. It's not malicious – we humans simply lack access to the messy, subconscious forces that actually drive our decisions. And the validation founders think they've achieved? Often just a mirage that feels good but offers little predictive power about market success. Things get worse when founders unknowingly structure interviews in ways that guide responses toward foregone conclusions. Ask someone about pain points in their category management process, and suddenly they'll discover frustrations they never knew they had. Inquire about automated solutions, and watch as enthusiasm materializes even from people who routinely ignore or disable automation in their current tools. Each question plants subtle suggestions about what should matter, unconsciously shaping responses that feel genuine but actually just mirror the conversation's structure. Most dangerous is how these sessions create a false sense of forward momentum – founders leave convinced they've done proper validation work and have permission to forge ahead with their existing vision, building for what customers say rather than what they actually do.
The fix isn't ditching customer research entirely but fundamentally shifting from asking to watching. When Dropbox existed only as a concept, Drew Houston didn't gather focus groups to chat about ideal file-sharing – he created a simple video demo and measured actual signup behavior. Those thousands joining the waitlist weren't politely expressing interest in an interview setting but taking concrete action that revealed genuine intent. This behavioral focus exposes the revealing gaps between what people say versus what they do. Observation uncovers the weird workarounds people have cobbled together, the friction points they've learned to tolerate, and the real problems they solve daily without even recognizing them as problems. Someone might never think to mention they take screenshots of important information instead of using bookmarks – not because they're hiding anything but because they don't consider it noteworthy. Yet that unconscious habit reveals volumes about their actual information management needs in ways no interview question could ever uncover.
The fix isn't ditching customer research entirely but fundamentally shifting from asking to watching. When Dropbox existed only as a concept, Drew Houston didn't gather focus groups to chat about ideal file-sharing – he created a simple video demo and measured actual signup behavior. Those thousands joining the waitlist weren't politely expressing interest in an interview setting but taking concrete action that revealed genuine intent. This behavioral focus exposes the revealing gaps between what people say versus what they do. Observation uncovers the weird workarounds people have cobbled together, the friction points they've learned to tolerate, and the real problems they solve daily without even recognizing them as problems. Someone might never think to mention they take screenshots of important information instead of using bookmarks – not because they're hiding anything but because they don't consider it noteworthy. Yet that unconscious habit reveals volumes about their actual information management needs in ways no interview question could ever uncover.
For founders building early products, this means transforming your approach: studying existing behaviors in natural settings, testing with tangible prototypes rather than hypothetical scenarios, measuring what people actually engage with rather than what generates enthusiastic nodding, and examining product failures with clinical detachment rather than through the distorting lens of post-rationalization. It means embracing that your potential customers aren't deliberately misleading you – they're misleading themselves, constructing narratives that make sense to them but have little relationship to how they'll behave when making real decisions. This gap between stated preferences and observed behavior isn't some minor inconvenience – it's precisely where countless promising startups go to die, building elegant solutions to problems people claim to have but never actually attempt to solve. The founders who break through aren't necessarily the ones who listen most attentively in interviews but those who observe most carefully in the wild – who design for how humans actually behave rather than how they wish to be perceived. In a market where most rely on the false comfort of interview validation, the real edge belongs to those willing to shut up and watch, to trust the evidence of behavior over the seduction of well-articulated needs, and to build for the messy humans revealed through observation rather than the idealized versions that show up to customer interviews.
For founders building early products, this means transforming your approach: studying existing behaviors in natural settings, testing with tangible prototypes rather than hypothetical scenarios, measuring what people actually engage with rather than what generates enthusiastic nodding, and examining product failures with clinical detachment rather than through the distorting lens of post-rationalization. It means embracing that your potential customers aren't deliberately misleading you – they're misleading themselves, constructing narratives that make sense to them but have little relationship to how they'll behave when making real decisions. This gap between stated preferences and observed behavior isn't some minor inconvenience – it's precisely where countless promising startups go to die, building elegant solutions to problems people claim to have but never actually attempt to solve. The founders who break through aren't necessarily the ones who listen most attentively in interviews but those who observe most carefully in the wild – who design for how humans actually behave rather than how they wish to be perceived. In a market where most rely on the false comfort of interview validation, the real edge belongs to those willing to shut up and watch, to trust the evidence of behavior over the seduction of well-articulated needs, and to build for the messy humans revealed through observation rather than the idealized versions that show up to customer interviews.
Stop Listening To Your Customers—Watch Them Instead
ESSAY
ESSAY
April 23rd, 2025
April 23rd, 2025
ESSAY
ESSAY
April 23rd, 2025
April 23rd, 2025
Stop Listening To Your Customers—Watch Them Instead
SARAH LEANS FORWARD eagerly, notebook in hand. "What features would you want in an app that helps you manage your finances?" she asks the potential customer sitting across from her at the coffee shop. The woman thinks for a moment, then launches into a wish list that sounds suspiciously like every finance app already on the market, plus a few impossibly complex capabilities that would require an army of engineers.
"I'd want it to automatically categorize my expenses, show me where I could save money, and maybe predict my future spending," the woman says confidently. "Oh, and it should be super simple to use."
Sarah nods enthusiastically, adding bullet points to her growing list of "validated" features. Her startup idea is taking shape through these customer interviews – twelve so far – each yielding similar responses that confirm what she already believed. By the end of the week, her product roadmap is bursting with features that customers claim they want. Six months later, her beautifully designed finance app launches to... silence. The very same people who eagerly described their perfect app don't download it, or download it and never return.
This scenario plays out daily across startup land, where the mantra "talk to your customers" has become sacred wisdom. The problem isn't that founders aren't listening – it's that they're listening to the wrong thing. Customer interviews are performances, not revelations. They capture what people think they should want, not what they actually need. The fundamental flaw in customer interviews is that humans are unreliable narrators of their own behavior. We construct stories about ourselves that bear little resemblance to our actual decision-making. When asked what we want in a product, we respond with rational-sounding features that make us appear thoughtful, forward-thinking, and sophisticated. We claim to value simplicity while demanding complexity. We insist we'll pay premium prices for quality, then choose the cheaper option when no one's watching. Early-stage founders face a particularly dangerous trap: the validation illusion. Customer interviews create an appealing mirage of market research while obscuring the actual terrain. Even worse, the very act of structuring interview questions inevitably shapes the responses. Ask about pain points in financial planning, and suddenly everyone has them. Ask about excitement for a new solution, and enthusiasm materializes out of thin air. The most damaging aspect of this illusion is that it feels like progress—founders walk away from these sessions with the comforting sensation of having done their homework, having validated their idea. This false validation becomes a subtle form of permission to proceed without genuine insight.
What's the alternative? Stop listening to your customers and start watching them instead.
When Dropbox was still just an idea, Drew Houston created a simple video demonstrating the product. He didn't ask potential users what file-sharing features they wanted; he showed them what was possible and watched their reactions. The video generated thousands of sign-ups for a product that didn't even exist yet. Those sign-ups represented a genuine behavioral signal – people taking action – rather than the polite nods of interview subjects.
Observation reveals the gap between stated intentions and actual behavior. It exposes the workarounds people have created, the friction points they've learned to tolerate, and the real problems they solve daily without articulating them as "problems." A person might never mention that they take screenshots of important information rather than bookmarking it – not because they're hiding it, but because they don't consider it noteworthy. Yet that behavior reveals volumes about how they actually manage information.
For early-stage founders, this means shifting from asking to observing:
Watching existing behaviors reveals uncomfortable truths about your market. Before building anything, study how people currently solve the problem you're targeting. The clumsy, inefficient solutions they've cobbled together contain the raw intelligence you need. A founder building a productivity tool might discover users don't actually want more efficiency—they want the appearance of efficiency to signal their value to others. This insight would never emerge from a direct question. Testing with prototypes instead of questions creates immediate behavioral evidence. Put something tangible in front of potential users—even a paper mockup—and watch what they do with it, not what they say. Their hesitations, the features they ignore, the workarounds they immediately attempt—these reveal the gap between stated desires and actual needs. When you do build something, measure clicks, not compliments. A user who spends four minutes in your app but claims to love it in an interview is sending contradictory signals—trust the behavior, even when it contradicts your hope. Most critically, study the failures with clinical detachment. When users abandon your product, resist the urge to ask them why directly. Instead, examine the last actions they took before churning. The digital breadcrumbs they leave reveal more than their post-hoc rationalizations, which are designed to protect their self-image rather than illuminate your product's weaknesses.
The hard truth for founders is that customer interviews feel productive while revealing almost nothing. They create an illusion of progress—pages of notes and enthusiastic feedback—while masking the reality that only behavior predicts behavior. Your customers aren't lying to you; they're lying to themselves, constructing narratives that make sense to them but bear little relationship to how they'll act when no one's watching. This gap between stated preferences and actual behavior isn't just a minor inconvenience—it's the graveyard where most startups are buried, having built perfect solutions to problems people claimed to have but never actually tried to solve. The uncomfortable wisdom here is that empathy itself can be dangerous when it's directed at what people say rather than what they do. True customer insight comes from the painful, messy business of watching humans bumble through their actual lives, making irrational choices, abandoning optimal solutions, and consistently failing to act in their own best interests. Want to build something people actually use? Accept that your customers are unreliable witnesses to their own experiences, then design for their actions, not their explanations. The founders who win aren't those who listen most attentively—they're the ones who watch most carefully, even when what they see contradicts everything they've been told.
ELLIOT CRANE