“I don't trust what people say they'll do; I trust what they actually do when asked to commit. Build a landing page, describe the product as if it exists, and run real traffic to a sign-up or pre-pay button. If strangers won't click or commit, fix the idea before you write code.”

“Before Superhuman, I would have said you validate an idea by talking to users and building an MVP. Now I think that's only half true.”

“I wasted months on products people said they'd use, then didn't. Now I validate by selling the promise before writing code: landing page, real traffic, and a commitment that hurts if they flake.”

“Post-PMF, I validate the next idea the same way I wish I had pre-PMF: thirty structured customer conversations before a single line of code, then a pre-sale or concierge MVP to make people pay. Enthusiasm is not validation, a credit card is.”

“I wasted six months, a thousand hours, and fifty thousand dollars building a product no one wanted, so now I validate before I build. My test is simple: run a lean landing page experiment, ask people to prepay or put down a deposit, and do in-depth interviews about how they actually solve the problem today.”

“I validate by putting money on the line fast: build the worst MVP in under a month, spend $5K on 10 to 20 influencer promos, and see if anyone pays. For Cal AI, the first prototype took three weeks, and if it hadn't worked immediately it never would have.”

“I wasted years treating validation as a vote: surveys, upvotes, friends saying that's cool. Post-PMF, I see it as a bank transfer. If a stranger won't click, sign up, or pay to solve the pain, you have theater, not traction. Run a smoke test, measure behavior, and let the market decide before you write a line of code.”

“Validation, from where I sit now, is not about the idea. It is about whether you can make people come back, pay, and tell a friend without you pushing them. I wasted years at one company building features for a customer hypothesis I had not confirmed; the real signal only appeared when I started interviewing the people who stopped using the product and found the shared reason they left.”

“I thought Airbnb was a silly idea in my YC batch, and Dropbox was obviously huge. I was wrong on both counts. So I stopped trusting my own taste and started asking whether customers pull the product, whether retention curves flatten, and whether the same kind of buyer keeps showing up for the same job.”