Write a Twitter/X thread

1/ Most startups don't die from competition. They die from building something nobody asked for — and 42% of failed startups cite "no market need" as the reason. Here's how to know before you waste 18 months.

2/ Talk to 30 customers before you write a line of code. Not "would you use this?" — that's a lie generator. Ask "what did you do last time you hit this problem?" Past behavior is the only honest signal.

3/ Charge before you build. A founder I know put up a $49 pre-order page for a tool that didn't exist. 12 people paid. That's not validation theater — that's 12 humans who'd be angry if it failed.

4/ Watch what they do, not what they say. A fintech team swore users wanted budgeting charts. Session replays showed everyone clicking "export to CSV" instead. They killed the charts, shipped CSV, retention doubled.

5/ Solve a problem people are already hacking around. If your users have a janky spreadsheet workaround, that's gold — the pain is proven, they just need something better.

6/ Beware the "vitamin vs. painkiller" trap. Vitamins are nice-to-haves people forget to renew. Painkillers get bought at 2am. Slack was a painkiller. Most failed apps were vitamins.

7/ Validation isn't a phase you finish — it's a habit you keep. The teams that win never stop asking "what did you do last time?" Bookmark this for your next idea, and follow for more brutal startup math.