Stewart Butterfield’s team spent years and real money building a game called Glitch. It flopped. What survived the wreckage was the internal chat tool they’d cobbled together to coordinate while building it, and that leftover tool became Slack. The company didn’t set out to build workplace software. It set out to build a game, noticed what people actually kept using, and rebuilt itself around the accident.
That’s the real shape of finding the thing customers want, and it’s a less flattering story than the one most pitch decks tell. The plan wasn’t validated by research; it was discarded by evidence. Nobody at Tiny Speck sat in a room and reasoned their way to a chat product. They shipped something, watched the data contradict the thesis, and had the discipline to believe the data over the thesis.
That discipline is rarer than it sounds, because founders today have more instrumentation than any generation before them. Dashboards refresh in real time. You can watch churn happen at the click level, segment it by cohort, trace it back to a specific release. It looks like control. It isn’t. Measuring a thing precisely and controlling it are different skills, and the gap between them is where most startups actually die. You can chart the storm in perfect resolution and still get soaked, because the chart never learned how to stop rain. The market moves on its own logic, not on your roadmap’s, and no amount of resolution changes that. What the dashboard gives you is better information about a system you still don’t run.
So what actually separates the founders who catch the signal from the ones who drown in their own metrics isn’t more data. It’s fewer bets. The startups that make it through the early years almost always describe the same move in hindsight: they stopped spreading time and money across everything defensible and narrowed hard, usually down to two or three things they were willing to be wrong about in public. Not because focus is a virtue in the abstract, but because a scattered company can’t tell the difference between a real signal and noise. When you’re doing twelve things at once, every metric looks like it’s trending somewhere. When you’re doing three, you actually know what moved and why.
Customer hostility is where this gets tested in real time, and it’s the part founders are worst at. Someone posts publicly that your product wasted their week, or your support team, or your billing, or the thing you shipped that broke their workflow the day before a deadline. The instinct is defensive: explain, minimize, wait it out. The better move, and the harder one, is to treat the hostile customer as a live data point instead of a threat to manage. Reply constructively. Fix the actual thing. Do it in public if the complaint was public. You cannot control whether someone likes you, and trying to is its own kind of instrumentation trap, the customer-relations version of thinking a chart controls weather. What you can control is whether the next hundred people who read the exchange see a company that responds like an adult or one that disappears into a support queue.
There’s a version of this that shows up in investing too. Every retail platform now hands you real-time price feeds, sentiment scores, options flow, alerts down to the second. It’s genuinely useful information, and it’s also a trap for the same reason the analytics dashboard is a trap: precision about a system is not leverage over it. The market doesn’t submit to your view of it just because your view is well-instrumented. The traders who do fine long-term tend to look almost boring by comparison, holding a short list of convictions and updating slowly, not because they see less than everyone else but because they’ve stopped mistaking visibility for control.
None of this is an argument against tools. It’s an argument about what tools are for. Slack’s founders didn’t need less data about Glitch to find the exit; they needed the willingness to let the data overrule the plan. The founders who survive customer anger aren’t the ones with better crisis playbooks, they’re the ones who stopped needing the customer to be wrong. And the investors who last aren’t the ones staring hardest at the feed, they’re the ones who decided in advance how few things they’d actually act on.
The instrumentation always improves. The humility to be corrected by it doesn’t come standard, and it’s the only part of the stack that was ever actually load-bearing.
Leave a Reply