"Pivot" is one of the most romanticized words in startup vocabulary. Most pivots are actually founders refusing to walk away from sunk cost. There's a real distinction between a useful pivot and a face-saving one.
What a real pivot looks like
A real pivot keeps one of: the audience, the channel, or the underlying insight. It changes the other two. Slack pivoted from a game (Glitch) — but kept the team chat tool they'd built for themselves. The insight survived; the product didn't.
What a fake pivot looks like
- "Same idea, different niche" with no new evidence the new niche wants it.
- "Same idea, AI-powered" with no new value created.
- "Same idea, B2B instead of B2C" because consumer didn't work.
Each of these is the same dead horse with a new hat.
The quit checklist
Quit when 4+ are true:
- You've done 30+ user interviews and the pain isn't there.
- A landing page test got under 1% conversion across 3 audience variants.
- The few users you have churn within 30 days.
- You've stopped enjoying the work entirely (not just hard days — months).
- You'd be relieved if a friend told you to stop.
The pivot checklist
Pivot when 3+ are true:
- One user segment loves the product and 3 others ignore it.
- Users repeatedly ask for a different but related thing you could ship in 2 weeks.
- A specific feature gets 10x more usage than the rest.
- You're acquiring users but they're not the ones you targeted.
Each is a real signal pointing somewhere specific.
The honest move when neither applies
If you're not getting clear quit signals or clear pivot signals, you probably haven't run enough experiments yet. Keep testing — see 48-hour validation experiments.
Quitting cleanly is a skill
Founders who quit cleanly almost always ship a winner within 12 months. Founders who pivot repeatedly often spend 3 years on the same dead idea wearing different hats. See Kill Your Darling Ideas.