The pivot mythology: Slack was a gaming company. Instagram was a location-sharing app. YouTube was a dating site. The right pivot at the right time created category-defining companies.
The reality: for every celebrated pivot, there are hundreds of founders who pivoted away from a working idea because growth was slower than they expected, or who stayed with a failing idea too long because they'd invested so much.
The decision framework for stay vs. pivot:
Persist when: you have genuine customer demand at small scale, you understand why growth is slow (distribution, not product fit), and you have a specific hypothesis about what would change the trajectory. Slow growth with clear product-market fit in a subset of your target market is a distribution problem, not a product problem.
Pivot when: the customers who use your product don't look like the customers you can economically acquire at scale, the most enthusiastic users aren't willing to pay what the unit economics require, or the problem you're solving keeps getting redefined by customer conversations in a direction away from your current product.
The hardest case: mixed signals. Some customers love it. Churn is moderate. Growth is slow. The product needs work but it's not broken. This is where most founders spend years.
The diagnostic for mixed signal situations:
Take your three most enthusiastic customers, your three customers with the lowest satisfaction, and three churned customers. Interview all nine. Do the three enthusiastic ones share characteristics that the others don't? If yes, you may have PMF in a smaller ICP than you're targeting. If the enthusiastic and unenthusiastic customers look similar, the problem is in the product.
Pivots work when they follow the signal. Follow the signal, not the mythology.