A pivot is a structured change in strategy based on what you have learned. It is not giving up. It is taking the evidence seriously and applying it. But knowing when to pivot versus when to persevere is genuinely hard. Pivoting too early means abandoning something that just needs more time. Pivoting too late means spending months or years on something the data already told you was not working. Ries proposes a regular review meeting, which he calls a pivot or persevere meeting, where the team examines the evidence and asks specific questions: Are our key metrics improving? If so, by how much? Is the improvement driven by our recent changes, or by something outside our control? Have we learned something new that changes our core assumption? The key discipline is separating the decision from emotion. Founders love their ideas. That love makes them persist long past the point when the data has spoken clearly. The structured review forces the data to be the deciding vote rather than attachment to the original vision. Ries catalogues specific types of pivot: changing the customer segment, the revenue model, the feature set, or the underlying technology. Each is a different kind of strategic change, but all share the same quality: they are driven by learning, not despair.