Series: Cognitive Biases
- The Danger of Cognitive Bias
- Action Bias
- Sunk Cost Bias (this post)
Sunk cost bias is one of the easiest traps to fall into because it feels like a virtue.
It sounds like commitment. It sounds like resilience. It sounds like backing your team and sticking with a plan. In creative work, especially, there’s a strong cultural pull towards pushing through. Finishing what you started. Not quitting.
The trouble is that sunk cost bias is not commitment to outcomes. It’s commitment to history.
It’s the tendency to keep investing in a direction because of what you have already put into it, rather than because it is the best choice now. Time, money, pride, reputation, internal politics. Once you have paid those costs, walking away feels like admitting the spend was wasted. So you keep going.
And by keeping going, you waste more.
Why it’s particularly dangerous in creative projects
Sunk cost bias thrives in games and software for a few reasons:
- Progress is ambiguous. You can’t always tell whether something is “nearly there” or fundamentally wrong.
- Effort is visible. Everyone remembers the late nights, the crunch, the sacrifices. That makes change feel ungrateful.
- Identity gets involved. Teams fall in love with solutions, not just problems. A direction becomes who we are.
- The end line moves. Features expand, expectations rise, and “done” becomes a negotiation rather than a state.
It is completely normal to feel attached to work you have invested in. That attachment is not the problem. The problem is letting that attachment drive decisions.
What it looks like in the room
Sunk cost bias usually sounds reasonable, which is why it is so hard to catch. It often comes out as:
- “We’ve come this far.”
- “It would be a shame to throw it away.”
- “We just need a bit more time.”
- “If we stop now it’ll all have been for nothing.”
- “We can’t change direction after what we’ve told everyone.”
Sometimes these statements are true. Sometimes continuing really is the best call. But if the main justification is past investment rather than present evidence, you are in danger.
A useful test is this:
If we had not started yet, would we start this today?
If the honest answer is no, continuing is probably sunk cost bias.
The costs
Sunk cost bias drains more than money.
It drains attention
The longer you invest, the more focus the direction consumes. It crowds out alternatives. You stop exploring because you are busy defending.
It drains morale
Teams can sense when they’re stuck polishing something that isn’t working. People lose faith. Cynicism builds. Good people leave.
It drains optionality
The more you commit, the harder it becomes to pivot. Work creates dependencies. Dependencies create schedules. Schedules create politics. Politics creates inertia.
It drains learning
If you can’t stop, you can’t admit something was wrong. If you can’t admit something was wrong, you can’t learn cleanly. You repeat the pattern.
How to counter it
You don’t eliminate sunk cost bias. You build a system that makes it easier to escape.
Create fixed evaluation points
Set regular checkpoints where the explicit goal is to evaluate direction, not just progress. Treat them as decision moments, not status updates.
Ask:
- What did we learn since the last checkpoint?
- What assumptions have been validated or invalidated?
- What is the next smallest test that would de-risk this?
Define clear goals upfront
Before starting, define what success means in measurable terms. Not because numbers are magic, but because they make reality harder to negotiate with later.
If you can’t tell what “good” looks like, sunk cost bias will fill the gap with “more”.
Make quitting respectable
Teams cling to bad directions partly because stopping feels like failure. So change the meaning. Treat a smart pivot as competence. Reward the person who spots the dead end early.
A practical tactic: keep a lightweight “kill list” of ideas you decided not to pursue, with a sentence on why. It normalises the act of stopping.
Separate effort from value
This is cultural. Praise effort, but do not confuse it with correctness. Make it normal to say “that was hard work, and it still isn’t the right approach”.
Seek external perspective early
People close to the work are the least able to see it clearly. An external producer, advisor, client, or peer can spot the gap between story and evidence.
The earlier you do this, the cheaper it is.
Prototype and iterate
Find the simplest possible way to express the idea, then test it. Paper prototypes, ugly builds, thin vertical slices. Anything that produces learning quickly and cheaply reduces the chance you’ll end up defending a large investment.
Budget mindfully
Recognise when additional investment is likely to produce diminishing returns. If the next month of work only moves you from “maybe” to “still maybe”, you’re probably feeding sunk cost bias.
A blunt but useful question:
What would we do with this time and money if this project did not exist?
Celebrate the learning
This isn’t motivational fluff. It’s a mechanism. If the only acceptable narrative is “we shipped and it was a win”, people will keep shipping the wrong thing to avoid the alternative narrative.
Make learning an explicit output. Write it down. Share it. Use it. That is how you convert sunk costs into future value.
Conclusion
Sunk cost bias is one of the most common causes of waste in creative work because it disguises itself as loyalty and grit.
The cure is not being cold or ruthless. The cure is being honest about evidence, building checkpoints that make re-evaluation normal, and creating a culture where stopping is sometimes the smartest form of progress.
If you’d like an external viewpoint on a project that feels like it’s being carried by history rather than signal, do get in touch.
