Sunk Costs

Overview

A sunk cost is money, time, or effort already spent that cannot be recovered. The sunk cost fallacy is our irrational tendency to continue investing in something because we've already invested heavily, even when abandoning it is clearly the better choice. Rational decision-making requires focusing exclusively on future costs and benefits, treating past investments as irrelevant. First formalized by behavioral economists Amos Tversky and Daniel Kahneman in 1972, this principle helps cut losses early, avoid throwing good money after bad, and make clear-eyed strategic pivots.

The concept appears across domains: businesses shutting down failing projects, individuals leaving unfulfilling careers, investors selling losing positions, and governments canceling wasteful programs. The key insight is that once resources are spent, they're gone regardless of your next decision. Your only question should be: "Given where I am now, what path creates the most value going forward?"

When to Use

Project evaluation: Deciding whether to continue or cancel initiatives that have consumed significant resources
Strategic pivots: Determining when to abandon an established strategy despite prior investment
Product development: Choosing whether to ship, iterate, or kill features after months of work
Career decisions: Evaluating whether to persist in a field despite years of training
Relationship assessment: Deciding whether to stay or leave long-term commitments
Investment management: Selling losing positions despite substantial unrealized losses

The Process

Step 1: Identify All Resources Already Spent

List everything invested to date: money, time, effort, reputation, opportunity cost. Label these explicitly as "sunk" - they're gone regardless of what you do next.

Example - Failed product launch:

$500K in development costs (sunk)
18 months of team time (sunk)
Delayed competitor launch (sunk opportunity cost)
Founder reputation tied to success (partially sunk)

Step 2: Frame the Decision From Zero

Ask: "If I were encountering this situation fresh today, with no prior investment, would I choose to start this path?" This mental reset removes emotional attachment to past decisions.

Reframe test: "A competitor just offered me this exact position in the project for free. Would I accept, knowing what I know now about likely outcomes?"

If the answer is no, you're likely being influenced by sunk costs rather than making the optimal forward-looking decision.

Step 3: Calculate Future-Only Cost-Benefit Analysis

Evaluate only what you'll invest going forward versus expected future returns. Past investments are irrelevant to this calculation.

Framework:

Future costs: Additional money, time, and opportunity cost required to continue
Future benefits: Realistic expected value from completion
Alternative use: What could you gain by reallocating these future resources elsewhere?

Example - Struggling startup:

Future costs: $200K more capital needed, 12 months more runway, team morale risk
Future benefits: 10% chance of $5M outcome = $500K expected value
Alternative: Pivoting to new idea with 40% chance of $2M outcome = $800K expected value
Decision: Shut down current path despite $500K already spent

Step 4: Set Pre-Commitment Kill Criteria

Before starting any project, define objective metrics that will trigger reconsideration, removing future emotional decision-making.

Examples of kill criteria:

"If we don't have 1,000 users after 6 months, we evaluate shutdown"
"If Series A fundraising fails after 25 investor conversations, we pivot or wind down"
"If customer acquisition cost doesn't drop below $100 by Q3, we stop spending on this channel"
"If we miss three consecutive quarterly revenue targets, we revisit the strategy"

Step 5: Schedule Regular Strategic Checkpoints

Create fixed intervals for reassessment independent of sunk costs. Ask: "Would we start this today?" at each checkpoint.

Checkpoint cadence:

Weekly: High-risk, fast-moving initiatives
Monthly: Standard product development
Quarterly: Strategic business units
Annually: Major infrastructure investments

Step 6: Use The "Monkeys and Pedestals" Test

Identify the hardest, riskiest part of your plan first. Test that assumption before accumulating sunk costs on easier supporting work.

Approach: Instead of building easy wins first, ask "What's the one thing that would make this entire effort pointless if it doesn't work?" Validate that first.

Example: Don't build a beautiful product interface before validating that customers will pay for the core value proposition. Otherwise, you'll have sunk design and engineering costs influencing you to launch something fundamentally non-viable.

Step 7: Adopt the "Revolving Door" Mental Model

Imagine you just walked into this situation as a new person without history. What would that person with fresh eyes and no emotional attachment decide?

Prompt: "If I quit tomorrow and a replacement was hired, what would they do given only the current situation?" Often this reveals that continuing is irrational, but you're trapped by sunk cost psychology.

Example: Netflix vs. Blockbuster

Blockbuster's sunk cost trap (2000):

Sunk: $5 billion in real estate, retail infrastructure, brand built on physical stores
Future cost: Transitioning to streaming would cannibalize existing business
Decision: Rejected acquiring Netflix for $50M, continued investing in stores
Outcome: Bankruptcy in 2010, Netflix worth $150B today

What rational analysis required: Ignoring the $5B sunk into stores and asking "If we started today, would we build a rental store chain or streaming platform?" The answer was obvious to outsiders, but insiders couldn't psychologically write off their prior investment.

Netflix's clear-eyed pivot: Even Netflix faced this decision. In 2011, they had sunk huge costs into DVD-by-mail logistics and brand positioning. When streaming became viable, they aggressively shifted despite short-term pain, because they correctly assessed future-only economics favored streaming. Revenue fell 25%, stock dropped 75%, but they ignored sunk DVD infrastructure and committed to the future.

Anti-Patterns

"We've come too far to quit now": Distance traveled is sunk. Only remaining distance matters.

"We can't let that money go to waste": The money is already wasted if continuing doesn't create value. Continuing wastes additional resources.

"Just a little more will get us there": The "just a little more" fallacy compounds sunk costs indefinitely. Set clear kill criteria upfront.

"But I've invested five years in this career": Past time is gone whether you stay or pivot. Only question: what's best use of your next five years?

"We need to at least ship something to show for the effort": Shipping bad products damages reputation more than nothing. Sunk costs don't justify adding future harm.

"The team will be demoralized if we cancel": Demoralization compounds if you continue a doomed project. Better to cut losses and redeploy talent.

Related Frameworks

Opportunity Cost: What you give up by continuing rather than pivoting to next-best alternative
Expected Value Thinking: Calculate probability-weighted outcomes to overcome emotional attachment
Pre-Mortems: Identify likely failure modes before accumulating sunk costs
Commitment Devices: Lock in kill criteria before emotions cloud judgment
Time Value of Money: Future resources are worth more than waiting to validate sunk investments
Loss Aversion: The psychological pain of realizing losses drives sunk cost fallacy
Creative Destruction: Willingness to abandon old investments for superior new paths