At a leadership event, a successful founder describes persistence as the reason the company survived. A manager takes the lesson back to a struggling project and argues for another funding round. The story may contain useful experience, but the decision requires information the founder’s account cannot supply on its own.
What happened to comparable projects whose leaders also persisted? What changed during the founder’s repeated attempts? How much additional funding was available, and what evidence suggested that another attempt was worthwhile? These questions help distinguish a plausible contribution to one outcome from a general rule for future decisions.
Success quotes compress experience into a memorable statement. They can express values or encourage effort without functioning as causal evidence. Problems arise when an organisation uses that compression to justify a strategy while leaving the selection of examples, relevant conditions, and possible costs unexplored.
Identify the Claim Behind the Advice
Advice about persistence can mean several things. It might encourage someone to tolerate an uncomfortable learning period, continue a viable strategy after an ordinary setback, or keep investing despite evidence that the underlying case has failed. Those are different recommendations.
Specify the behaviour and intended outcome before evaluating the story. For the struggling project, the question might be whether funding one further experiment is likely to produce decision-relevant evidence at an acceptable cost. That is more precise than deciding whether the team has enough determination.
Separate a value from a prediction. A person can value courage or craftsmanship regardless of commercial success. A claim that a behaviour increases the probability of a particular outcome needs evidence about that relationship.
The account may still help generate a hypothesis. The founder may explain a useful change in customer research or a way of managing cash during uncertainty. Treating that detail as something to investigate preserves more of its value than accepting or dismissing the whole story.
Ask How the Examples Were Selected
A conference programme, business book, or list of admired companies is a selected sample. Organisations that closed, remained small, or never attracted attention may be absent. Their absence can change the relationship that appears in the visible cases.
Denrell’s analysis of undersampling failure shows how observing surviving organisations can produce misleading beliefs about management practices. The analysis concerns a selection mechanism; it does not establish that every practice used by a successful company is ineffective.
Consider a deliberately hypothetical set of 100 ventures:
| Approach recorded at the outset | Ventures | Successful outcomes |
|---|---|---|
| Persistence-focused strategy | 90 | 18 |
| Other strategy | 10 | 2 |
Among the 20 successes, 18 used the persistence-focused strategy. Looking only at them could make persistence seem decisive. In the full example, both groups have a success rate of 20%. The prevalence of a behaviour among winners does not by itself show that it raised the chance of winning.
Real comparisons are harder because strategies, circumstances, and definitions vary. The example illustrates the missing denominator, not a finding about entrepreneurship. It shows why the manager needs information about comparable attempts before using a selected story to justify further funding.
Distinguish Selection, Attribution, and Outcome Bias
Survivorship bias concerns which cases enter the evidence. Attribution concerns how causes are assigned within a case. Outcome bias concerns how knowledge of a result changes the evaluation of the earlier decision. These can interact without being the same error.
A founder may accurately report hard work while underestimating the contribution of distribution access, timing, or favourable competition. An observer may focus on personal qualities because they are visible in the story. Neither possibility establishes that effort did not matter.
In Ross, Amabile, and Steinmetz’s social-role experiment, assigned roles influenced how knowledgeability was judged. The study illustrates the need to account for advantages created by a situation. It does not permit a complete explanation of any particular person’s success through circumstances alone.
Baron and Hershey’s outcome-bias experiments examined how outcomes affected evaluations of decisions. For a project review, the practical implication is to inspect the information and reasoning available when the decision was made, alongside the eventual result.
A profitable gamble can have been poorly justified, and a reasonable decision under uncertainty can end badly. The result is important evidence, but one result does not reveal the quality of the process without further context.
Reconstruct the Mechanism and the Conditions
Useful advice should explain how the proposed behaviour could affect the outcome. Persistence might create more opportunities to learn, but those opportunities matter only if the attempts produce information and the organisation can change its approach.
For the struggling project, examine what the next round would do differently. Would it test a new customer segment, remove a known constraint, or repeat the same launch with a larger budget? A general statement about endurance cannot distinguish these options.
Identify resources and protections. A founder able to survive several failed attempts may have funding, skills, access, or alternative income that another team lacks. These conditions affect the consequences of following the advice. They are part of the case, not reasons to disregard the person’s contribution.
Also examine what success means. Revenue growth, profitability, social impact, personal satisfaction, and survival are different outcomes. A story selected for one may provide limited guidance for another. The manager should not inherit an objective simply because it made the original story compelling.
The forecasting article explains how to make assumptions and comparisons explicit. The same discipline helps identify which part of a success story might reasonably travel to a new setting.
Look for Evidence Beyond the Most Visible Survivor
Seek comparable cases with a range of outcomes, including attempts that stopped. Define the group in a way relevant to the decision: similar resources, market conditions, stage, and strategy. Comparing a small internal experiment with a mature market leader may create more apparent insight than useful evidence.
Records made before outcomes are known can help. Decision notes, budgets, customer findings, and dated forecasts show what people expected and what they changed. Retrospective accounts can still contribute, but their explanations should be checked against those records where possible.
Define the observation period as well. A strategy can appear successful if the account ends just after a favourable result, while its later costs remain outside the story. Conversely, a difficult early period can conceal benefits that require time to appear. For the project, establish when the outcome will be judged and include the resources consumed along the way. This avoids comparing one company’s early difficulties with another’s eventual success as though the cases were observed on the same basis.
Do not assume that failure stories are automatically more truthful. They can also simplify causes, defend reputations, or reflect unusual conditions. Including them improves coverage of the evidence; it does not remove the need to examine the account.
Where comparative data are limited, state the uncertainty. A bounded experiment may provide a better next step than adopting a broad principle throughout the organisation. The test should specify what would count as useful evidence and how much the organisation is prepared to spend obtaining it.
Verify the Words Separately From the Lesson
When publishing an attributed quote, check a primary source such as the original speech, interview, book edition, or letter. A quote collection repeating a famous name is not a substitute for establishing where the wording appeared.
Record whether the text is exact, translated, shortened, or paraphrased. Check the surrounding argument as well. A remark about persevering with practice may become misleading when presented as advice never to close an unviable business.
If the source cannot be established, omit the attribution or explain the idea in your own words without presenting it as a verified quotation. An appealing line does not need a famous author to be discussed, but the reader should be able to distinguish a documented statement from a generic example.
Authenticity and validity remain separate. A verified quote can contain weak advice, while an unattributed observation can suggest a useful question. Establishing who said something does not establish that following it will improve the present decision.
Turn the Advice Into a Reviewable Decision
For the manager considering more funding, translate persistence into a proposal with a purpose, cost limit, evidence requirement, and review point. State what the additional work is expected to reveal and what would lead the organisation to continue, change direction, or stop.
Keep the decision conditions visible before the outcome. This makes it easier to learn from the result without rewriting the original rationale. It also reduces the chance that continued spending becomes a test of commitment rather than an assessment of the project’s prospects.
The temporal-discounting article examines another influence on that choice: current pressures and delayed consequences. A team might abandon worthwhile work too quickly or continue unpromising work for reasons a success slogan cannot diagnose.
The founder’s experience can remain part of the discussion. Its strongest contribution may be a specific practice, a warning about a difficult stage, or a question the team had overlooked. Once the organisation examines comparable cases and the mechanism involved, it can use that experience without asking a memorable success story to carry more evidence than it contains.
