# Strategies for Learning from Failure

*Edmondson · Safety & Error, Team Learning · Harvard Business Review · 2011 · Paywalled*

Diagnoses why so few organisations learn from failure despite near-universal commitment to doing so: not a lack of effort, but managers thinking about failure the wrong way, treating it as uniformly bad and learning from it as simple (reflect, exhort, write a report). Opens with the blame game as the core obstacle — failure and fault are near-inseparable in most cultures, and executives Edmondson surveyed estimate that only 2–5% of organisational failures are genuinely blameworthy, yet 70–90% get treated as if they were, which is why so many failures go unreported and their lessons lost. A 'spectrum of reasons for failure' running from deliberate deviance (clearly blameworthy) through inattention (context-dependent — a fatigued worker on an overlong shift is less at fault than whoever scheduled it) to thoughtful experimentation (potentially praiseworthy) is offered as the corrective to blanket blame. The paper's most durable contribution is a three-part taxonomy of organisational failure that has since become one of the most widely reproduced frameworks in the management literature: preventable failures in predictable operations (routine, high-volume work where deviation is genuinely bad, addressed by checklists and Toyota's andon cord); unavoidable failures in complex systems (inherently uncertain work — ER triage, aircraft carriers, nuclear plants — where small process failures are inevitable, and most serious accidents result from chains of small unnoticed ones lining up); and intelligent failures at the frontier, a term borrowed from Duke's Sim Sitkin for failures that generate genuinely new knowledge because the right answer wasn't knowable in advance — 'trial and error' is called a misnomer here, since 'error' implies a correct answer existed to be missed. IDEO's small, deliberately unannounced pilot of a new strategic-innovation service with a mattress-company client, which failed at its specific goal but taught the firm what to change, is the illustrating case: strategic innovation services went on to become more than a third of IDEO's revenue. Frames organisational learning around three activities — detection, analysis, experimentation — each anchored by a memorable case. Detection: Alan Mulally's colour-coded status reports at Ford, where every manager reported green despite billions in losses until Mulally applauded the first honest yellow report and broke a costly silence; and NASA's Columbia disaster, where a rigid, schedule-obsessed hierarchy made it hard for engineers to raise anything short of a rock-solid concern about a piece of foam strike, and the ambiguity went unresolved for 16 days until the fatal re-entry. Analysis: complicated by fundamental attribution error (downplaying our own responsibility for failure while blaming others' character for theirs), and illustrated by a 2010 New England Journal of Medicine finding that North Carolina hospitals hadn't become measurably safer after over a decade of heightened awareness of medical error, against the positive counter-example of Intermountain Healthcare's systematic analysis of physicians' protocol deviations. Experimentation: contrasts basic-science researchers' comfort with high failure rates (70% or more in some fields) against business pilots typically designed under optimal rather than representative conditions — illustrated by a pseudonymised telecom company's DSL launch, which followed a successful small pilot staffed by expert reps and tech-savvy customers straight into a full launch that missed 75% of its commitments, because the pilot never tested the unrepresentative, lower-skill conditions of the real rollout. Eli Lilly's 'failure parties', held since the early 1990s to honour intelligent experiments that didn't pan out, are offered as a low-cost way to reduce the stigma that keeps failing projects funded long after the data says they shouldn't be. Concludes by reframing the standard leadership worry, that tolerating failure produces a lax, mistake-prone culture, as a false choice: organisations that catch, correct and learn from failure before their competitors do will win; those still playing the blame game will not.

- **This page:** https://explore.psychsafety.com/n/edmondson-2011/
- **View the source paper:** https://hbr.org/2011/04/strategies-for-learning-from-failure
- **Interactive map:** https://explore.psychsafety.com/?mode=papers&node=edmondson-2011

## Connected concepts (10)

- [Psychological Safety and Learning Behavior in Work Teams](https://explore.psychsafety.com/n/edmondson-1999.md) (paper)
- [The Fearless Organization: Creating Psychological Safety in the Workplace for Learning, Innovation, and Growth](https://explore.psychsafety.com/n/edmondson-2018.md) (paper)
- [Behind Human Error: Cognitive Systems, Computers, and Hindsight](https://explore.psychsafety.com/n/woods-et-al-1994.md) (paper)
- [How Complex Systems Fail: Being a Short Treatise on the Nature of Failure, How Failure is Evaluated, How Failure is Attributed to Proximate Cause, and the Resulting New Understanding of Patient Safety](https://explore.psychsafety.com/n/cook-1998.md) (paper)
- [Normal Accidents: Living with High-Risk Technologies](https://explore.psychsafety.com/n/perrow-1984.md) (paper)
- [Out of the Crisis](https://explore.psychsafety.com/n/deming-1986.md) (paper)
- [The Tenerife Disaster](https://explore.psychsafety.com/n/the-tenerife-disaster-of-1977-were-going.md)
- [Blametropism](https://explore.psychsafety.com/n/blametropism.md)
- [Learning from Error](https://explore.psychsafety.com/n/learning-from-error-or-punishing-it.md)
- [Watermelon Effect](https://explore.psychsafety.com/n/the-watermelon-effect-and-greenwashing.md)