Knowing When to Pivot or Stop

There is a point on many journeys where the road you are on no longer leads where you hoped. Recognising it is hard. Persistence is celebrated, and every story of success includes a stretch of doubt that was overcome. Yet persistence has its failures too, and these are told much less often. The founder who kept going, long after the evidence had turned, has no conference talk.

This guide is about that moment. It does not tell you to quit or to continue. It offers ways to look at the evidence, to separate the feelings from the facts and to choose among three honest options: persist, pivot or stop. It also covers how to carry each out with care for the people who depend on you.

The three options

Persist. Keep going on the same path, perhaps with changes of tactics. This is right when the evidence shows progress, however slow, and when you have good reason to think more time will help.

Pivot. Change direction in a structured way. A pivot, in the language of lean product development, is a change of strategy without a change of vision, based on what has been learned. You might change the audience, the problem you solve, the way you reach people or the business model, while keeping the insight that matters.

Stop. Close the product, deliberately. This may mean winding it down, selling it, giving it away or handing it to a community.

None is a failure by itself. The failure is to drift: to continue without a decision, spending time and goodwill on a path that you have not examined.

Why this is hard

Several habits of mind make the decision difficult.

Sunk cost. The sunk cost fallacy is the tendency to continue an endeavour because of the effort and money already invested, rather than judging it by future prospects. The past is gone either way. Only what comes next matters to the decision.

Survivorship bias. We hear from the people who persisted and succeeded, and not from those who persisted and did not. The stories give a distorted picture of the odds.

Identity. After months or years, a product can feel like part of you. Closing or changing it can feel like a judgement on yourself.

Sporadic encouragement. A kind email, a good week or a small mention can feel like a sign, when it is noise.

Fear of the alternative. Not knowing what comes next makes the present path seem safer than it is.

Naming these makes them easier to resist.

Look at the evidence

Gather facts, not feelings. Over a meaningful period, such as the last three to six months:

• Your main measure. Has it moved in the direction you want? By how much? Is the trend steady or accidental?
• Retention. Is there a group of users who stay and a curve that flattens?
• Pull. Do users contact you unprompted, refer others and ask for more?
• Attempts. How many well-designed experiments have you run? What did they show?
• Learning. Are you still finding things out that change your view, or are you repeating the same tests?
• Resources. How long can you continue, in money, energy and the goodwill of the people around you?
• Opportunity. What else could you do with that time?

If you have run several honest, one-at-a-time experiments and the main measure has not moved, that is information. If you have not run honest experiments, you may not yet know.

Signs that persisting is reasonable

• The curve of retention shows a floor above zero.
• One group of users clearly values the product, even if they are few.
• Recent changes have moved numbers in the right direction.
• You have specific, untested ideas with a reasoned chance of working.
• The need is real and growing, and competitors, if any, are not solving it.
• You can sustain the work for long enough to test further.
• You still believe in the problem, and you are learning.

Signs that a pivot may be right

• A small group of users has strong pull, but it is not the group you built for.
• Users keep using one part of the product and ignore the rest.
• The problem is real, but the way you reach people is not working.
• The product works, but the market you aimed at cannot pay or is too small.
• Users solve a different problem with your product than you intended.
• You have learned something that makes the original plan no longer sensible, but not the vision.

A pivot keeps the insight and changes the approach. Examples include narrowing to a segment, switching channel, changing the business model or focusing on the feature that users like and dropping the rest. The best pivots are built on evidence from users, not on boredom.

Signs that stopping may be right

• Repeated, well-designed attempts have not moved the main measure.
• No group shows strong pull, and conversations reveal indifference, not frustration.
• The need is weaker than you believed.
• The resources are running out, without a realistic route to more.
• You no longer believe in the problem, or in your ability to solve it.
• The cost, including to your health and relationships, outweighs the plausible gain.
• A better use of your time is clear.

Stopping is easier to consider when you have a time limit set in advance: "If by the end of June we have not reached this milestone, we will wind down." Setting such a limit when you are calm prevents the endless drift.

A structured way to decide

1. Write down the evidence, in numbers and quotations, for and against.
2. Name the options with concrete descriptions: persist with these changes, pivot to this, stop in this way.
3. State what would have to be true for each option to succeed. Which of those assumptions can you test cheaply?
4. Set a deadline for the decision, and a short test where useful.
5. Consult someone you trust who is not emotionally invested, and ask them to challenge you.
6. Decide, and write the decision and its reasons.
7. Tell the people affected, promptly and honestly.

The act of writing makes a difference. It separates the case from the mood.

Pivoting well

If you decide to pivot, do it deliberately.

• Keep what you learned. Write down the insight that remains valid.
• Form a new hypothesis, and a milestone to test it.
• Talk to the group you are pivoting toward before building.
• Tell existing users what is changing and what it means for them.
• Honour commitments you have made, and give notice if something will no longer be supported.
• Update public pages so that they match the new direction. Old claims should not linger.
• Be willing to relaunch, at a new stage, with an honest account of the change.

Do not pivot to escape a hard problem. If the difficulty is shared by the new direction, you will meet it again.

Stopping well

Closing a product is an act with duties.

1. Decide the date, and give users enough notice for them to plan, as a matter of respect. Check your terms and any commitments.
2. Tell people plainly, with the reason in a few words and what happens next.
3. Let users take their data. Provide an export, or help them move.
4. Honour what you owe, including any paid commitments, in line with your terms and the law.
5. Keep the site useful. A short page explaining the closure, rather than a broken link.
6. Secure what remains. Remove access, delete data that you should not keep and shut down services that could become security risks.
7. Thank people. Users, collaborators, supporters.
8. Record what you learned, privately and perhaps publicly.
9. Consider alternatives: a sale, an open-source release or a handover to someone who will care for it, where this serves users.
10. Look after yourself. Take time to rest before the next thing.

A well-handled closure leaves a good reputation behind, and many founders find that it is the basis for the next project.

The conversation with co-founders and team

Disagreement is likely. One person may be ready to stop and another ready to push on.

• Share the evidence first, before opinions.
• Say what you feel and what you fear, openly.
• Agree a method, such as a time-boxed test, before agreeing the outcome.
• Respect the commitment each person has made. Equity, income, reputation and personal circumstances differ.
• Plan for parting well, if you disagree.

Many partnerships survive a difficult decision made honestly. Few survive one made through silence.

A worked example

A pair of founders has run a meal-kit-planning product for fourteen months. Sign-ups have plateaued, retention flattens at a low level and several experiments on onboarding and pricing have not moved the figure. They write down the evidence. Overall retention is low. But a segment, people managing diets for medical reasons, retains at more than double the rate, writes unprompted emails and has twice referred friends. Everyone else treats the product as a curiosity.

They consider three options. Persist unchanged: unlikely to work, given a year of data. Stop: tempting, but the segment's pull is the strongest signal they have ever seen. Pivot: narrow the product to dietary management, rebuild the first-use path around it and talk to twenty people in that group before building anything.

They set a milestone and a deadline: ten weekly active users in the new segment within ten weeks, or they will wind the product down with a month's notice. They talk to the segment, rewrite the page and tell existing users what is changing. At week ten, they have fourteen. They continue, and the next milestone is clear. Had the number been three, they had agreed to stop, and had written down how they would do it. The deadline made either decision easier.

Questions makers ask

Is it always best to try longer? No. More time is valuable only if you are learning and the evidence points to a plausible route.

How do I avoid giving up too soon? Run honest, disciplined experiments before concluding. A pile of untested ideas means you have not yet tested enough.

What if I am not sure? Choose a short, specific test and a deadline. Uncertainty is a reason for a plan, not for drift.

Will stopping harm my reputation? Rarely, if you handle it honestly. People remember how you treated them.

Summary

At a fork, you can persist, pivot or stop. The difficulty comes from sunk cost, survivorship bias, identity and fear. Look at the evidence over time: the main measure, retention, pull, attempts, learning, resources and opportunity. Persist when you are learning and progress is plausible, pivot when a small group's pull points elsewhere and stop when disciplined attempts have not moved the needle. Decide in writing, set deadlines, treat users fairly and take care of yourself and your partners. A clean decision, however hard, is better than a long drift.