The Escalation Trap: When Sunk Costs Block Your Next Decision

The moment you decide to keep investing in a failing strategy because you've already invested heavily in it, you've entered a decision trap that behavioural science has documented for decades—yet most organisations still walk into it with eyes open.

Sunk cost fallacy operates with particular force in strategic contexts because it masquerades as prudence. A marketing campaign underperforms. A product pivot hasn't gained traction. A vendor relationship has become costly and misaligned. The rational response is to cut losses and redirect resources. Instead, decision-makers often double down, justifying continued investment by referencing what's already been spent. The logic feels sound: abandoning the initiative now would mean "wasting" all previous expenditure. But this reasoning inverts the actual decision problem. Past spending is irrelevant to whether future spending will generate returns.

What makes this particularly insidious in organisational settings is that sunk cost reasoning often wears the costume of commitment and follow-through. Leaders who escalate commitment are frequently praised for persistence. They're seen as having conviction. The narrative becomes one of resilience rather than cognitive bias. This social reinforcement makes the trap deeper.

The mechanism works like this: once you've publicly committed to a direction, abandoning it creates cognitive dissonance. You've staked reputation on the decision. Colleagues have aligned their work around it. Stakeholders have been told it's the priority. Reversing course now feels like admitting error—not just in judgment, but in character. So instead of treating the new information (the strategy isn't working) as a reason to change course, you treat it as a reason to prove the original decision was right. You invest more to "give it a fair chance." You reframe the metrics. You extend the timeline. You find reasons why the current data is misleading.

The cost of this trap extends beyond the wasted capital. It's an opportunity cost that compounds. Resources committed to a failing initiative are resources unavailable for initiatives that might actually work. Teams become demoralised by being asked to rescue something that should have been abandoned. Decision-making credibility erodes when leaders are seen to be throwing good money after bad, even if they frame it differently internally.

The distinction between sunk cost fallacy and legitimate persistence is worth examining closely. Legitimate persistence involves new information that actually supports continued investment—market conditions have shifted favourably, a technical breakthrough has emerged, competitive dynamics have changed. Sunk cost escalation, by contrast, involves no new positive information. It's driven entirely by the desire to justify past decisions.

One practical signal: if your primary argument for continued investment is "we've already spent X," you're in the trap. If your argument is "new evidence suggests Y," you might be making a sound decision. The difference is whether you're reasoning forward or backward.

The harder question is structural: how do organisations create conditions where sunk cost reasoning is less likely to dominate? Some possibilities emerge from decision science research. Separating the decision to continue from the decision to have started helps. Assigning a different decision-maker to evaluate continuation than the one who initiated the strategy reduces ego investment. Setting predetermined decision points—"we'll evaluate this on date X against metric Y"—creates friction against drift. Building psychological safety around course correction means leaders aren't punished for changing direction when evidence warrants it.

The escalation trap is particularly dangerous because it doesn't feel like a trap while you're in it. It feels like prudent stewardship of resources, like seeing something through, like having backbone. Only in retrospect, when the full cost is visible, does the pattern become clear. By then, the damage is done.

The question isn't whether your organisation will face this decision. It's whether you'll recognise it when it arrives.