How to stop past investment from deciding your next move

Review a struggling project by comparing the future consequences of the available options, including genuine obligations and transition costs. Keep irrecoverable spending in the historical record without treating it as a benefit of continuing. Ask whether you would choose the next commitment today on the evidence now available.

The sunk-cost effect is an increased tendency to continue an endeavor because money, effort, or time has already been invested when that prior investment should not determine the current choice. Arkes and Blumer’s 1985 paper defines and examines the effect. Their research is a foundation for the decision question here, not proof that a particular executive or project is affected by it.

Find the argument that only looks backward

Composite scene: You open a project proposal at home on a Sunday morning. The first page emphasizes what the company has already invested. The new work is described as finishing what you started. The document gives little attention to what the same people could do elsewhere or what stopping would actually require.

Ask the sponsor to explain the next commitment without relying on the prior spending as its justification. The history may explain why the company is in the situation. It does not by itself establish that more work is the best available use of resources.

Do not conclude that stopping is automatically right. A project with a difficult history may still have the best prospective case. A project near completion may still have unfavorable future consequences. The review must establish the actual options and relevant evidence.

Separate sunk costs from future switching costs

A sunk cost is a past cost that the current choice cannot recover. A future switching cost changes what happens if the company chooses a different path. They belong in different parts of the review.

In a hypothetical software implementation, completed work may be irrecoverable. A remaining contractual obligation, migration requirement, or customer transition consequence may still affect the next choice. Qualified finance, legal, and operational owners must verify how those items apply in the actual case.

Do not label every unpleasant cost sunk in order to simplify the argument. Likewise, do not place past expenditure in a future-benefit column because management dislikes describing it as waste. Keep the classification inspectable and use qualified advice where accounting, valuation, or contractual questions arise.

The fresh-entry review

1. Reconstruct the current position

Ask the relevant owners to identify work completed, assets or capabilities now available, remaining obligations, and decisions still open. Establish which amounts or resources are truly unaffected by the next choice.

Use permitted records rather than a sponsor’s headline summary. A useful existing capability should not disappear from the analysis merely because its creation cost is sunk. The question is what the company now has and what each option would require from this point onward.

2. Ask the fresh-entry question

Use: “If we inherited this position today, with the current evidence and obligations, would we authorize the proposed next step?” Then ask for the reason.

The question removes neither obligations nor knowledge acquired through the project. It changes the burden of explanation. A recommendation should identify prospective value and consequence, rather than rely on the fact that management previously backed the work.

3. Compare feasible future options

Include continuing, changing scope, pausing for defined information, or stopping where those are genuinely available. Ask what each option means for customers, capacity, commitments, and the intended business outcome.

Do not compare a fully developed continuation plan with a vague instruction to stop. Give the alternative enough operational detail to expose its consequences. Equally, do not bury continuing costs inside the existing budget while showing every cost of changing direction separately.

4. Verify transition consequences

Have qualified owners establish contractual duties, recoverable assets, customer obligations, and the work needed to change course. Distinguish confirmed requirements from assumptions about what a supplier or stakeholder might accept.

A transition may create meaningful future costs. Include them honestly. The fresh-entry test is not permission to ignore them or to provide informal financial advice. Where a decision affects investment, financing, employment, or binding commitments, obtain the appropriate professional review.

5. Authorize the next step on its current case

The proper owner should state the chosen option, evidence, remaining uncertainty, and review trigger. If continuation is justified, identify what makes the next commitment worthwhile now. If stopping is justified, specify the authorized transition rather than announcing a symbolic reversal.

Keep the original history and the current decision separate. A changed recommendation does not require pretending the earlier choice was obviously wrong. It requires explaining why the present choice follows the evidence and obligations now available.

Use an option comparison with a common starting point

A working comparison can show the action from today, required future resources, expected outcome with stated uncertainty, transition requirements, unresolved information, and authority needed. Use the same time period and scope for each feasible option.

For the hypothetical implementation, continuing may require a remaining integration task. Changing scope may preserve one useful capability while removing another. Stopping may require a customer handoff or contractual process. Have the actual owners establish the relevant details before the company commits.

Do not create precision the evidence cannot support. If expected benefits remain uncertain, show the assumption and its source. Ask qualified analysts to perform any necessary financial evaluation rather than assigning a convenient number during the leadership discussion.

Examine the waste argument without diagnosing the sponsor

Arkes and Blumer discuss evidence linking continued commitment with concerns about wastefulness. Their paper does not establish that every person defending a project is motivated by that concern. Ask what the sponsor means when they say stopping would waste the prior investment.

They may be identifying an asset that can still produce value, a real future obligation, or a historical expenditure that cannot change. Clarify which. The answer determines whether the argument belongs in the prospective comparison.

Use proposed wording:

I want the history recorded accurately. For the next choice, separate what cannot be recovered from what we would gain, spend, or owe under each option. Explain the continuation case as if we were evaluating the current position for the first time. We will include real transition consequences rather than treating stopping as costless.

Keep later evidence distinct from the original research

Zhu and Feldman’s 2025 registered replication revisits Arkes’s later wastefulness research and reports mixed results across its scenarios. It is not a direct replication of every experiment in the 1985 paper.

The distinction matters when explaining the evidence to an executive team. Do not say the whole field is settled because a later study supports part of a mechanism. Do not say the sunk-cost effect is disproven because a particular scenario did not replicate. State which finding the source actually examines.

The advisory review here does not depend on assigning a diagnosis of bias. Even without knowing why someone prefers continuation, management can require a defensible comparison of future options.

Preserve what the team knew at each decision

Fischhoff’s 1975 research examines the effect of outcome knowledge on judgments about earlier events. Keep a dated record of the assumptions and information behind the original project approval and the current review.

A later failure should not erase earlier uncertainty. A later success should not transform weak evidence into a strong original process. Ask which facts were available, which could reasonably have been obtained, and which arrived only afterward.

Industrial/organizational psychology, the psychology of work and organizations, helps examine these judgment and accountability questions. The fresh-entry method remains a management aid, not a validated score of decision quality or a substitute for financial analysis.

Separate accountability from defending the old choice

The sponsor may need to explain a previous commitment. Give that review an appropriate process. Do not make continuation the only way to avoid personal embarrassment, and do not make stopping a required public admission of incompetence.

Likewise, do not use a fresh-entry review to excuse concealed evidence or misconduct. If the record raises such concerns, use the proper formal route. The prospective choice and the accountability review can inform one another without being collapsed into a single verdict about a person.

If you sponsored the project yourself, state that involvement. Ask a suitable reviewer to challenge the current option comparison. Retain the responsibility to decide within your authority while making your own historical commitment visible.

Include the work the project displaces

Ask the operating owners what the same scarce capacity could support under each feasible option. Compare genuine alternatives, with the same care used to assess continuation. Do not invent a more attractive project merely to defeat the existing one.

If no alternative can use a resource immediately, record that condition. It may affect the current decision without making historical expenditure a future benefit. A useful comparison preserves both the available opportunity and the practical limits on moving people or commitments.

Check whether the next commitment follows the new rationale

After the authorized decision, inspect the actual instruction. If the company continues, confirm that scope and resources match the reviewed case. If it changes direction, confirm that the transition owners understand their obligations.

Watch for an unapproved return to the old plan through routine spending or private instructions. A decision to stop is incomplete when no one owns the remaining work. A decision to continue is incomplete when the conditions that justified it have already changed.

The useful question remains prospective: given today’s evidence, available capabilities, and genuine obligations, what should the company authorize next? Answer it without requiring the past to look better or worse than the record supports.