Sunk-Cost Commitment
Continuing a failing bind because past investment is treated as reason to persist — backward-looking, not forward
Decisions must be revisitable on qualified evidence and forward-looking cost — past spend alone must not block supersession when triggers fire or analysis warrants reversal.
Sunk-cost commitment is the failure where an organisation continues a committed direction primarily because too much was already invested — money, reputation, careers, narrative — not because forward-looking analysis shows staying course beats reversing from here. “We’ve come too far to turn back” replaces expected-cost comparison.
Past spend is sunk; it cannot be recovered by persistence. The relevant question is marginal: given where we are now, does reversal or continuation minimise expected organisational loss? Personal loss — face, standing, career capital for the owner and allies — is also real, but must not be smuggled in as if it were organisational switching cost; that conflation is forced alignment and defensiveness dressed as prudence (We Already Decided, case 4).
In software systems
Legacy platform migration continues after load tests fail because “we’ve already rewritten three services.” Technical debt project absorbs budget despite falsified performance assumptions — no superseding ADR because admission is politically costly. Teams ship the wrong abstraction because refactor “would waste the last two quarters.”
In human organisations
Acquisition integration proceeds after due-diligence assumptions fail — exit would embarrass leadership. Market entry persists in a dead segment because launch spend is on the books. Headcount and roadmap stay locked to a strategy deck nobody would choose today if starting fresh.
Often disguised as credible commitment theatre — or mislabeled credible commitment device — but without time bounds, tier honesty, and declared end date in the (A)DR, it is denial with strategy vocabulary.
In socio-technical systems
Vendor contracts and multi-year licences lock direction after early signals fail; switching cost is real but inflated by narrative. Automated spend gates continue funding a model path because training cost is cited — without tiered forward analysis or trigger review.
Structural causes
Consensus culture
Public reversal humiliates the room that nodded at bind. Sunk cost becomes shared face to protect.
Counter direction
Record forward-looking reversal analysis at bind and on trigger review. Separate execution runway (bounded hold to finish work) from indefinite persistence. Supersede when triggers fire regardless of spend already sunk — We Already Decided, cases 2 and 4.