Sunk cost

A sunk cost is a cost that has already been incurred and cannot be recovered, regardless of future decisions. The money is spent, the time is gone, the effort is complete - nothing you do now changes that. Rational decision-making should ignore sunk costs and focus only on future costs and benefits. Yet humans consistently fall into the "sunk cost fallacy," letting past investments inappropriately influence future choices.

Why it matters

Product teams make sunk cost errors constantly. We continue features that aren't working because we've invested so much. We persist with technologies that slow us down because switching feels wasteful. We ship products that shouldn't launch because cancelling feels like admitting failure.

Understanding sunk costs matters because it enables rational decisions that focus on future value rather than past investment. It supports better pivots by killing failing initiatives when evidence warrants. It prevents escalation of commitment from throwing good money after bad. It reduces emotional attachment by separating past effort from future decisions. And it improves resource allocation by directing resources to best opportunities, not past commitments.

The sunk cost fallacy

The sunk cost fallacy is the tendency to continue an endeavor because of previously invested resources rather than future expected returns. We've spent six months on this feature, so we should ship it - even if evidence now shows nobody wants it. We've invested $1 million in this technology, so we should keep using it - even if a better alternative has emerged. We've been in this market for years, so we should keep trying - even if it's not working.

This fallacy feels rational in the moment. "We've already invested so much" seems like a good reason to continue. But the investment is gone regardless of what you do next. The only question is: given where you are now, what's the best path forward?

Sunk costs in product development

Feature investment represents months spent on a feature that user research now shows won't be used. The sunk cost fallacy says ship it anyway because we've invested so much. The rational response is to kill it and use resources on something valuable.

Technology choices include legacy systems that slow down development, but switching would require significant effort. The sunk cost fallacy says keep using the legacy system because we've already learned it. The rational response is to compare future costs of continuing versus switching, ignoring past investment.

Market persistence means we've been in this market for three years without traction. The sunk cost fallacy says persist because we've invested so much. The rational response is to evaluate future potential objectively - if it's not working, pivot.

Team investment includes people we've trained who aren't performing or aren't a fit. The sunk cost fallacy says keep them because we've invested in development. The rational response is to assess future value, not past investment.

Product continuation applies to products that aren't meeting expectations. The sunk cost fallacy says continue because killing it wastes previous investment. The rational response is to evaluate future potential honestly.

Recognizing sunk cost thinking

Watch for sunk cost fallacy signals in your thinking and conversations: "We've already spent..." as a reason to continue. "It would be a waste to stop now." "We've come too far to turn back." "Think of all the work that would be lost." "We need to see this through." These phrases indicate sunk cost thinking. The question should be: what's the best use of future resources, regardless of past investment?

Making rational decisions

Ignore what you've spent because past costs are irrelevant to future decisions. What matters is the future cost versus future benefit.

Compare future paths by asking what it costs to continue versus what benefit you expect. What does it cost to stop versus what does it free up for other use? What are the opportunity costs of continuing?

Seek outside perspective from people without emotional investment who can evaluate more objectively. Fresh eyes see sunk cost traps more clearly.

Pre-commit to evaluation by deciding in advance when and how you'll evaluate initiatives. "We'll ship the beta and assess after 30 days. If usage is below X, we'll cut it." Pre-commitment reduces the influence of sunk costs when the decision point arrives.

Celebrate good kills by rewarding the discipline to stop failing initiatives, not just the discipline to persist. Killing something that isn't working is a success, not a failure.

When sunk costs seem to matter

Some situations look like sunk costs but aren't quite. Completion costs sometimes require finishing past work to capture any value - a half-built bridge provides no value. But this is about future completion cost versus future benefit, not about the sunk cost of past construction.

Reputation effects exist because abandoning initiatives may damage credibility. But this is a future cost of abandonment, not a sunk cost. It should be weighed against future benefits of alternative paths.

Learning value means past investment may have generated knowledge useful for other pursuits. The knowledge isn't lost even if the project is killed. This knowledge is a sunk benefit, not a reason to continue.

Contractual obligations may require completing certain work regardless. These are future costs (of breaking contracts), not sunk costs.

Sunk costs and product culture

Organizations that handle sunk costs well create safe environments for honest evaluation without fear of blame, celebrate learning including from failed initiatives, make pivoting normal and not shameful, evaluate outcomes rather than just effort, and distinguish between poor decisions and poor luck.

Organizations that handle sunk costs poorly punish failure without distinguishing preventable from unpreventable, create strong attachment to initiatives, escalate commitment to justify past decisions, avoid killing anything once started, and confuse effort with value.

Sunk costs and product management

Product managers face sunk cost situations regularly in roadmap decisions (whether to continue initiatives that aren't performing), feature prioritization (when to cut versus continue), technical decisions (when to rebuild versus maintain), and market strategy (when to pivot versus persist).

Tools like Klero help by providing objective customer evidence. When you can see what users actually want and how they respond to what you've built, sunk cost decisions become more grounded in evidence rather than emotional attachment to past work.