Skip to content
Log in
← Glossary

sunk cost

Also called: sunk costs, sunk cost fallacy

Money already spent that no decision can bring back. It must not influence the choice about what to do next.

A sunk cost is money already spent that no decision can bring back: the development budget of a half-built feature, research that is already paid for, a prototype that will be thrown away. Because it is the same whatever the company decides next, it has no place in the decision.

The trap works in two directions. "We can't waste what we've spent" keeps a weak project alive (the sunk cost fallacy); "it will never pay back its cost" kills a project that is worth finishing. Both arguments look backwards. The right question looks forward: from today, what will each option cost and bring?

Sunk costs still matter for learning and accountability: a post-mortem should ask why 960,000 went into something that now looks weak. They also tend to sneak back into business cases as depreciation of past spending or as "investment to date" lines. Keep them in a separate note, outside the numbers the decision rests on.

Example

Klasna has spent 960,000 UAH on an exam simulator. Finishing it needs 150,000 more (an editor 90,000, a licence 60,000) and is expected to bring 380,000 of contribution: 380,000 − 150,000 = 230,000 from here on. The same developers could instead build package-renewal reminders worth 260,000.

The 960,000 is spent whichever way Klasna goes, so it enters neither option. "We can't waste 960,000" argues for the simulator; "230,000 will never pay back 960,000" argues against it. Neither is the reason: the reminders win by 260,000 − 230,000 = 30,000 because of what each option brings from now on.

The same applies to the 650,000 already spent on EU market research: it stays out of the funding request for the EU plan.

Common mistakes

  • "We've come too far to stop." How far you've come is sunk; only the remaining cost and value count.
  • "It will never earn back what it cost." Equally backward-looking; it can kill a project worth finishing.
  • Smuggling it in as depreciation. Writing off past spending over future years doesn't make it a future cost.
  • Adding it to a funding request. Ask for what the plan still needs, not what has been spent.
  • Ignoring it entirely. Leave it out of the decision, but review why it was spent.

Learn it in the course