The Sunk Cost Decision Framework: How to Make Better Choices in Business

The Problem

Your brain is broken when making decisions that involve past investments.

Your gut says: "Protect what you have invested."

Your brain should say: "Make the decision that creates the most value going forward."

They conflict. Your gut wins. You make bad decisions.

This framework fixes that.

The Simple Framework

Step 1: Identify the Decision

Write down what you are deciding:

"Should we continue with Product A or kill it?"

"Should I stay at this job or leave?"

"Should we rewrite this codebase or maintain it?"

"Should we keep this vendor or switch?"

Step 2: List Past Investments (Sunk Costs)

Write down everything you have invested:

Time: 18 months

Money: $1.5M

People: 5 engineers

Reputation: Told investors this would work

Opportunity cost: Did not pursue alternative

Important: Do not delete this list. Just acknowledge it.

Your brain wants to use this to make decisions. You are not going to let it.

Step 3: Cross Out All Sunk Costs

Physically draw a line through the list.

This is psychological. You are telling your brain: "This does not factor in."

Mantra: "That money is gone. That time is gone. It does not matter."

Step 4: Evaluate Current State

Ignoring the past, evaluate the present:

For Product A:

Monthly active users: 5,000

Monthly growth rate: 2%

Customer satisfaction: 6/10

Revenue: $50k/month

Burn rate: $80k/month (losing $30k/month)

Objective rating: 4/10 current health

Step 5: Project Future State (12 Months)

If you continue on the current path:

Optimistic:

Growth accelerates to 10%

Revenue reaches $150k/month

Profitability achieved

Realistic:

Growth stays 2%

Revenue stays $50k/month

Keep losing $30k/month

Runway exhausted in 6 months

Pessimistic:

Competitors dominate

Growth goes negative

Revenue drops to $30k/month

Runway exhausted in 3 months

Probability:

Optimistic: 10%

Realistic: 60%

Pessimistic: 30%

Expected value: (0.1 × 150k) + (0.6 × 50k) + (0.3 × -80k) = $15k + $30k - $24k = $21k/month (conditional on continuing)

But that assumes resources are "free." They are not.

Step 6: Evaluate Alternatives

Alternative 1: Pivot to adjacent market

Development cost: $200k

Timeline: 3 months

Success probability: 40%

Revenue if successful: $200k/month

Expected value: 0.4 × $200k = $80k/month

Alternative 2: Kill product, redeploy team

Redeploy 5 engineers to Product B

Product B success probability: 60%

Product B projected revenue: $150k/month

Expected value: 0.6 × $150k = $90k/month

Alternative 3: Sell company

Current valuation: $500k (based on $50k monthly revenue)

Offer: $300k (buyer sees realistic growth)

Expected value: $300k one-time

Step 7: Compare Expected Values

Option	Expected Value	Notes
Continue Product A	$21k/month	Declining
Pivot to new market	$80k/month	3-month risk
Redeploy to Product B	$90k/month	6-month ramp
Sell	$300k	One-time

Clear winner: Redeploy to Product B ($90k/month vs $21k/month)

Step 8: Check for Sunk Cost Bias

Ask yourself: "Is my recommendation influenced by the $1.5M invested in Product A?"

If yes, go back to Step 1.

If no, proceed.

Step 9: Make the Decision

Based on expected value (ignoring sunk costs):

Decision: Pivot or redeploy.

Not because of sunk costs. Because it is the rational choice.

Real Example: The Vendor Decision

Step 1: The Decision

"Should we switch from Vendor A (expensive, old-fashioned) to Vendor B (cheaper, modern)?"

Step 2: Sunk Costs with Vendor A

Contract cost paid (non-refundable): $200k

Integration effort: 3 months of engineering time ($100k)

Training investment: $50k

Historical data locked in: (High switching cost)

Total sunk: $350k+

Step 3: Cross It Out

That $350k is gone. Do not factor it in.

Step 4: Current State

Vendor A:

Monthly cost: $50k

Downtime per month: 4 hours (business impact: $20k)

Total cost: $70k/month

Vendor B:

Monthly cost: $20k

Estimated downtime: 1 hour (impact: $5k)

Total cost: $25k/month

Monthly savings: $45k

Step 5: Future Projection

Vendor A (stay):

Growing more expensive each year

Downtime likely increases with scale

Projected annual cost: $1M+ by year 3

Vendor B (switch):

Stable pricing

Better reliability at scale

Projected annual cost: $300k by year 3

3-year cost difference: $2.1M - $900k = $1.2M difference

Step 6: Switching Cost

One-time migration: $100k (engineering, data export)

Lost productivity during switch: $50k

Risk of data loss: $100k (probabilistic)

Total switching cost: $250k

Step 7: Analysis

Scenario	3-Year Cost	Notes
Stay with A	$2.1M	Growing
Switch to B	$900k + $250k (switch) = $1.15M	Cheaper overall

Savings: $950k over 3 years

Payback period: 6 months

Step 8: Check Bias

Is the $350k sunk cost making me want to stay?

No. The math clearly says switch.

Decision: Switch to Vendor B

Not because we want to waste the $350k sunken investment.

Because it is the right financial decision for the future.

Real Example: The Employee

Decision

"Should we keep this underperforming employee or replace them?"

Sunk Costs

Hiring cost: $20k

Training investment: $30k

Ramp-up time (productivity loss): $40k

Total: $90k

Current State

Current employee:

Salary: $120k/year

Productivity: 60% of ideal

Value delivered: $70k/year

Morale impact on team: Negative (worth -$20k/year)

Total value: $50k/year

Replacement cost:

Hiring: $20k

Training: $30k

Lost productivity (first 3 months): $30k

Total first year: $80k

Then: Salary $120k + full productivity ($130k) = $10k positive

Analysis

Keep underperformer:

Year 1: $50k value

Year 2: $50k value

Year 3: $50k value

3-year total: $150k

Replace:

Year 1: -$80k (switching cost)

Year 2: +$10k

Year 3: +$10k

3-year total: -$60k (better!)

Savings from replacement: $210k over 3 years

Decision: Replace the employee.

Not because you want to waste the $90k training investment.

Because it is better for the business long-term.

Quick Decision Checklist

When facing a decision:

[ ] What did I invest in the past? (Note it)

[ ] Is that investment gone? (Likely yes)

[ ] Does my recommendation rely on "I already invested"? (If yes, reconsider)

[ ] What is the expected value of each option going forward? (Calculate)

[ ] Which option creates the most future value? (Choose that)

[ ] Am I comfortable with this decision without mentioning past investment? (If yes, good)

Advanced: Multi-Factor Analysis

For complex decisions, consider:

1. Financial value (primary)

2. Strategic fit (secondary)

3. Team morale (tertiary)

4. Market opportunity (primary)

But do not factor in past investment.

When Others Use Sunk Cost Against You

Your boss says

"We invested $500k in that technology. We have to stick with it."

"I understand we invested $500k. But that cost is already incurred. The question is: what creates the most value going forward? Based on the analysis, switching to new technology saves $1M over 3 years. I recommend we prioritize future value over protecting past investment."

How to present it

Do not say: "We should ignore the $500k."

Say: "The $500k is a sunk cost. Let's focus on what generates the best returns going forward."

Make it about "future optimization" not "ignoring the past."

Conclusion

Sunk cost fallacy is the #1 reason businesses make bad decisions.

This framework eliminates it.

The process:

1. Acknowledge past investment

2. Cross it out mentally

3. Compare expected future value of options

4. Choose the best option

5. Do not mention past investment in your decision

If your decision requires you to cite past investment to justify it, the decision is probably wrong.

The best decisions are self-evidently good based on future value alone.