When to pivot

A pivot is a structured change in strategy without throwing away everything you learned — a new customer segment, a new job-to-be-done, a new revenue model, sometimes a new product wedge. It is not renaming the company every six weeks.

Pivot from evidence, with runway left. Do not pivot from boredom, embarrassment, or one loud investor.

The big idea

A sailor changes heading and keeps the logs. A flail throws the charts overboard and paints a new name on the hull.

The analogy breaks when nothing you learned transfers. Then you are starting over while calling it a pivot. Good pivots keep customer insight, technical leverage, and distribution relationships. They change who the customer is, what job you do, or how you charge.

Walking through it

Pivot when:

• You have evidence the current path will not reach a viable business (no retention, no willingness to pay, CAC permanently broken)
• You have a specific new hypothesis grounded in user conversations, not vibes from Twitter
• You still have enough runway to test the new hypothesis (often 6+ months after a cut)

Do not pivot when:

• You are bored or embarrassed
• One loud investor hates the category
• You have not talked to users in a month
• The "pivot" is actually seven products in a trench coat

Slack began as a gaming company (Tiny Speck / Glitch). The game failed commercially. The internal communication tool the team built for themselves had pull. That is a pivot with retained assets: a working collaboration product and a team that knew real-time systems.

Contrast with a team that rebrands from "AI for legal" to "AI for healthcare" to "AI for sales" every quarter without shipping depth. That is not pivoting. That is flailing.

Default alive vs default dead is the cash frame around the same decision. At current burn and growth, do you reach cash-flow break-even before money hits zero without a new raise? If not, you are default dead — every plan assumes investors will save you. Practical response while you still have time: cut burn, narrow to the segment that pays and retains, raise from a clear path, or choose an orderly shutdown while assets still have value.

Common mistakes

• Pivoting from boredom. The trough of sorrow is not evidence.
• A quarterly category hop with no depth. Flail.
• Pivoting with two months of runway. You cannot test a new hypothesis.
• Throwing away the only asset (insight, tech, or distribution) along with the failed packaging.

Check yourself

No retention, no willingness to pay, six months of runway after a cut, and a new hypothesis from user calls. Pivot?

Yes. Evidence, a specific hypothesis, and time to test it.

Why is Slack → chat a pivot and "AI for X" every quarter not?

Slack kept a working product and a team skill. The other keeps only a label.

Key takeaways

• A pivot keeps what you learned and changes customer, job, or model.
• Need evidence, a specific hypothesis, and runway — often 6+ months.
• Do not pivot from boredom, one investor, or a silent month.
• Default-dead companies that wait for a hot market are not pivoting. They are stalling.

Further learning

How Not to Die (Paul Graham)How Not to Diepaulgraham.com

Next: shutting down well.