Sunk cost
escalation of commitment
Money, time or effort already spent and not recoverable. Because it cannot be recovered, it should not affect what you do next — but it reliably does, because walking away feels like admitting the spend was wasted.
In practice
Two years into a product with no customers. The two years are the argument for year three, and they are the one fact that is identical whether you continue or stop.
The common mistake
Framing the decision as whether to waste the investment. It is already spent in both options, so the only question is what the next year is best used for.
A sunk cost is money, time or effort already spent that cannot be recovered. It should have no bearing on what you do next, and it reliably does.
The error
The sunk cost fallacy is continuing something because of what has already been invested. Two years into a product with no customers, the two years are the reason given for a third.
The logic is clear and the pull is strong. The only question that matters is whether continuing is the best use of resources from today forward. What was spent is gone in both branches, so it cannot distinguish between them.
Why it holds
Abandoning something requires admitting the earlier spending produced nothing, and that admission is the actual cost being avoided. Continuing defers it.
The pull is stronger when the commitment was public, when you chose it personally, and when the alternative is unclear. Organizations amplify all three, which is why projects survive long past the point where anyone privately believes in them.
There is also a real asymmetry in how the two errors feel. Stopping produces a visible, attributable loss on a specific date. Continuing produces a diffuse loss that never gets recorded as a decision. See hidden cost.
Getting out
Ask the fresh-start question. Knowing what you now know, would you begin this today? A no means the only remaining argument is the spending.
Make the alternative concrete. Stopping is hard when it means nothing; easier when it means starting something specific.
Set the exit before you enter. Decide in advance what evidence would mean this is not working, and what date you will check. A rule written before the commitment is not defending anything.
Separate the decision from the judgment of the person who made it. Most sunk cost persistence is reputation management, and treating a stop as information rather than failure removes the incentive to keep going. See loss aversion.
Concept web
Open the full webQuestions
What is the sunk cost fallacy?
Continuing an activity because of resources already invested rather than because of expected future returns. Past spending cannot be recovered, so it should not influence the decision.
How do you avoid the sunk cost fallacy?
Ask whether you would start this today knowing what you now know. Set exit criteria before committing, and treat stopping as information rather than as personal failure.
Why is the sunk cost fallacy so hard to resist?
Because stopping requires admitting the earlier investment produced nothing, which creates a visible loss on a specific date. Continuing defers that admission indefinitely.