Near Misses as Learning Signals

Definition

A near miss is an event that could have caused serious harm but didn't, due to chance, luck, or last-minute prevention. Unlike failures, which are visible and force investigation, near misses often go unexamined because they appear to be successes—the airplane landed safely, the product didn't fail in the field. Yet near misses reveal the same latent errors that cause catastrophes. Organizations that actively identify, investigate, and learn from near misses can correct root causes before they trigger disasters. Organizations that treat near misses as evidence of good management (normalization of deviance) drift toward catastrophe.

In the Book

Catherine Tinsley, Robin Dillon, and Peter Madsen describe how latent errors—flaws in process or design—compound invisibly until an "enabling condition" converts them into crises. They use three cases.

In 2009, Toyota had received 2,000+ complaints of unintended acceleration since 2001. Complaints were trending upward, but the company treated them as individual incidents rather than signals of a latent design error. When the Saylor family tragedy occurred—a Lexus accelerating uncontrollably and crashing—the same latent error suddenly became visible and forced a massive recall costing $2 billion and immeasurable reputational harm.

JetBlue Airways operated aggressively in bad weather, pulling planes away from gates to position them for takeoff. This risky strategy produced a years-long string of near misses—flights took off and landed successfully despite tight margins. But the strategy revealed a latent vulnerability. In February 2007, an ice storm created an enabling condition, and the near misses became a crisis: passengers stranded for 11 hours on tarmac, massive flight cancellations, and lost customer loyalty.

Apple's iPhone had an antenna design flaw that caused signal drop-off when held in a certain way. This flaw existed across all smartphones and had gone unremarked for years—each successful call was a near miss. When social media amplified one negative Consumer Reports review, the enabling condition was met, and a reputational crisis erupted.

The authors describe seven strategies to identify and respond to near misses. These include: learning from deviations by examining why acceptable-risk-thresholds have shifted; uncovering root causes rather than symptoms; demanding accountability (requiring managers to justify why they downgraded a near miss); considering worst-case scenarios explicitly; evaluating projects at every stage (not just postmortems of failures); and rewarding staff who report near misses rather than punishing them.

Why It Matters

In stable, low-consequence environments, near misses rarely matter. In complex, high-consequence systems—aviation, medicine, nuclear power, drug manufacturing—near misses are the only warning system available. By extension, in any organization pursuing ambitious goals with tight margins, near misses are where learning happens cheapest. An organization that normalizes deviance and dismisses near misses as "how we operate" is not learning; it's drifting toward the event that will finally force change. An organization that treats near misses as rich data can correct course before disasters force it to.