A feature interviews ten thriving cafés and finds that all their owners opened early. The headline suggests early opening is the secret to success. Before changing your own plans, ask a less glamorous question: what happened to the cafés that opened early and later closed?
This invented example illustrates survivorship bias: the visible survivors may be a selected part of the original group. The Catalogue of Bias’s explanation of selection bias describes the wider problem of drawing conclusions from a group that differs systematically from the population of interest.
Find the starting population
Look for who could have appeared in the story, not only who was interviewed. Were the examples chosen from current businesses, published authors, completed projects or people still responding to a survey? Each choice can leave out a different set of experiences.
Then ask what claim is being made. An interview can describe how one owner works. It does not, by itself, establish how often that method succeeds or whether it caused the outcome. A vivid personal account and a comparison of all relevant cases answer different questions.
Keep the useful detail without inventing a rule
You do not have to discard every success story. It may contain a practical idea worth testing on a small scale. Keep the claim modest: “This is something one surviving business does” is different from “Businesses survive because they do this”.
If the article includes data, look for the entry criteria, time period and what happened to cases that disappeared. If that information is absent, the honest conclusion may be that the story cannot answer the larger question.
A useful reading note separates the reported experience from the inferred lesson. Write down the observation, the missing comparison and what evidence would change your view. That small pause protects against turning the most visible outcomes into a misleading picture of everybody who tried.