Set stop-loss policies before starting projects

Define exit criteria at the start, when you are not yet sunk.

Why it works

The sunk cost fallacy is hardest to resist when you are already invested. Stop-loss policies are a precommitment device: by specifying exit criteria in advance, you make the decision at a time when no sunk costs exist to bias it. The pre-committed policy then serves as an anchor that resists the emotional pull of accumulating investment.

How to do it

1. Before starting any project with meaningful cost, write explicit stop criteria: "I will stop if X happens by date Y."
2. Specify the criteria in measurable, observable terms — not "if it’s not working" but "if revenue is below $N after six months."
3. Share the criteria with someone who will hold you to them.
4. When a trigger is hit, treat continuation as a fresh decision that must justify itself without reference to the prior investment.

Evidence

Precommitment devices have consistent empirical support across behavioral economics: committing to a rule before emotional states arise produces better outcomes than deciding in the moment. Ariely and Wertenbroch demonstrated this for deadlines; the principle extends to stop-loss rules. (observational)

Stop-loss policies only work if they are actually enforced; social accountability (telling someone) materially increases compliance.

Sources

• Ariely & Wertenbroch (2002), procrastination, deadlines, and performance, Psychological Science

Common mistake

Setting stop criteria that are vague enough to always be interpreted as "not quite triggered yet" — providing false comfort while still accumulating sunk cost.