How to Decide Without Chasing Sunk Costs

Set a new decision date

Imagine both options become available today. List future costs, benefits, and obligations from this point onward.

Set a review before the next meaningful commitment, not after it has already been spent. Freeze new discretionary work until the review is complete. Bring current evidence, remaining cost, expected future benefit, and the best alternative. A scheduled choice point prevents continuation from happening through a series of small approvals that never reconsider the whole project.

Keep real transition costs

Switching may require retraining, migration, cancellation, or relationship repair. These are future costs and belong in the comparison.

Keep future transition and shutdown costs in the comparison because they have not yet been spent. Include contract duties, data migration, customer support, staff time, and loss of useful service during change. Label them clearly so the team does not confuse a real cost of stopping with past spending that cannot be recovered.

Extract learning without buying more

Past effort may reveal useful information. Record the lesson separately so stopping does not feel like erasing the work.

Write the useful lesson from past work as a separate asset: a tested assumption, reusable component, trained skill, or documented failure mode. Then ask whether continuing is the best way to use that learning. Ending the current path does not erase the lesson, and buying more of the same is not required to prove that the earlier effort had value.

Use a continuation threshold

Define the future result that justifies another unit of time or money. Review at that point instead of renewing by default.

Define what the next investment must produce and by when. Include a stop condition if that result does not appear. Compare the cost of reaching that evidence point with the value it can create or clarify. A vague hope of getting closer to completion is not a continuation rule. The next step must earn approval from today's choices.

Use a scheduled continuation decision

Projects often continue because no moment is designated for stopping. Set a review before the next commitment of money or time. Bring the original success condition, current evidence, remaining cost, switching cost, and best alternative to that review.

The review is not a search for someone to blame for the past. Its job is to decide whether the next unit of investment is justified. Preserve a separate retrospective for accountability and learning so those needs do not distort the forward choice.

If stopping creates a visible write-off, compare that write-off with the cost of reaching the next meaningful evidence point. Do not fund a cosmetic milestone merely to make the project look complete. The next milestone should reduce uncertainty or produce usable value, not only postpone recognition of the loss.