By Flow State Leadership
Persist when evidence still supports the core customer problem and you have a credible, affordable next test. Consider a pivot when repeated evidence challenges a core assumption about the customer, offer or route to market. Decide with explicit criteria and constraints, not sunk costs or a single disappointing week.
General educational guidance. Illustrations are hypothetical, not client results.
Which assumption is failing?
Separate the problem, customer, offer, channel and delivery model. Low sales alone cannot tell you which one is wrong. Review actual customer behavior, delivery experience and cash obligations alongside what people said.
Ask whether the test was fair: Did the intended buyer see a clear offer? Could you deliver it? Was the buying cycle long enough to observe? A poorly executed test is not strong evidence that the problem does not exist.
What would persistence test?
Name one unresolved assumption, one proportionate test and the evidence you need. Set a decision date and a spending limit that respects existing obligations. Persistence should buy learning, not simply more activity.
In a hypothetical software business, interested managers never secure purchasing approval. A useful next test could involve the budget owner and a clearly scoped paid pilot. More demonstrations to the same non-buyers may teach very little.
When does a pivot deserve attention?
Consider changing direction when credible tests repeatedly challenge the assumption you depend on, or when delivery economics and available resources make the current path untenable. Explain which assumption changes and which evidence stays relevant.
In the software example, buyers might consistently value a different workflow than the one offered. A narrower offer could be worth testing. That is a hypothesis—not proof that the adjacent market will buy, and not a reason to rebuild the whole product immediately.
How do you compare the options?
Write a short decision note for persist, pivot and pause. For each, list supporting evidence, the strongest counterargument, cost, reversibility and the next decision point. Ask someone to challenge the preferred option without defending the work already spent.
Cash runway, contracts and obligations may limit the choices. Seek qualified financial or legal advice where those constraints matter. There is no universal threshold or checklist that guarantees the right answer.
How do you give the team direction?
Explain what is known, what remains uncertain, what changes now and what does not. Give the next test an owner and a review date. Avoid presenting a hypothesis as a settled strategic truth.
Adaptation is not directionless motion. Preparation makes it easier to notice new evidence and respond quickly while keeping the team clear about the work in front of them.