May 6, 2026

Most Startup Ideas Die From a Conversation the Founder Refuses to Have

The single most expensive habit in early-stage founders is not bad code, bad marketing, or premature scaling. It's the conversation they keep not having.

I've watched dozens of founders, including me on three separate occasions, sink six to twelve months into an idea they could have killed in a week if they'd just talked to the right five people. They didn't talk to those people. They knew, on some level, exactly what those people would say, and that's why they didn't talk to them. The whole arc of the project is then a slow, expensive way of avoiding the conversation.

This is what most startup idea validation content gets wrong. It tells you to "talk to customers," as if the problem is that founders don't know they should. They know. What they don't do is talk to the specific customers whose response would actually kill the idea — the ones with budget, with authority, with experience in the space. They talk to friends. They talk to other founders. They build a landing page. They post on Twitter. They get vague signal back and call it validation.

The conversation they refuse to have sounds something like this: Hi, I see you're the person who would buy this. Would you? And then, in the response — the wince, the polite redirection, the "interesting, send me a deck" — there is enough information to either go all-in or shut the project down.

You can usually feel which conversation you're avoiding. There's a specific person, or a specific kind of person, you keep almost-emailing and not. That is the conversation. Have it this week. If you can't bring yourself to have it, that is also information — your conviction in the idea isn't strong enough to survive a single honest signal, which means it definitely isn't strong enough to survive twelve months of building.

The other thing founder advice gets wrong is treating "kill your startup idea" like a worst-case outcome. It is not. Killing a bad idea in week two is a win. The cost is one weekend of work and a slightly bruised ego. The alternative — discovering in month nine that the idea was always going to die, after you've burned savings, optionality, and the patience of everyone around you — is several orders of magnitude more expensive.

I now run every idea through a simple test before I let myself build anything. Who is the person whose "no" would kill this? Have I talked to five of them? What did they say in the first sentence of their reaction, before they got polite?

If you can't answer those three questions, you don't have a startup. You have a project you're enjoying.

Get the brutal version of those answers from somewhere that won't be polite. Then either build it for real or shut it down by Friday.