The single biggest cost in early-stage building isn't time spent on the wrong product. It's time spent emotionally married to the wrong product. You can rebuild a codebase in a weekend. You can't easily walk away from an idea you've been telling people about for six months.
Why ideas are sticky
You told your spouse. You told your friends. You bought the domain. You sketched the logo. Every one of those small acts increases the cost of admitting it isn't working. By month three, you're defending the idea instead of evaluating it.
The pre-mortem
Before you start, write the failure obituary. "This idea died because ___." Be specific. Then ask: which of those failure modes can I cheaply prevent before committing? That's your validation work (see Should I Build This?).
The "kill criteria" rule
Before you start, write down the conditions under which you'd quit. "If after 30 days I have fewer than 3 paying users from my warm network, I stop." Future-you will rationalize ignoring those criteria. Past-you, writing them now, is more honest.
What to do when you kill an idea
- Don't pivot. Pivots are usually a way to keep the sunk cost alive. Walk away cleanly.
- Write a postmortem (see A Postmortem of 3 Apps That Scored A and Still Failed) so future-you doesn't repeat the same mistake.
- Take 1–2 weeks off ideas entirely. Outsiders specifically need this — your domain expertise will surface the next idea naturally.
The killing makes the next one possible
The founders who eventually ship a winner almost always have 1–3 dead ideas behind them. The dead ones aren't failures — they're tuition. The trap is staying on a dead idea long enough that you don't have the energy or runway for the next one. Don't fall for the vibe-coding trap of shipping forever instead of stopping.
If you suspect your current idea is the wrong one, take the ShiporDrop quiz with brutal honesty. A score under 40 isn't a verdict — it's permission to stop.