Most executives believe that failure is bad to learn from mistakes - just Ask people to reflect on what they did wrong and exhort them to avoid similar mistakes in the future
X
Failure is not always bad
Learning is not straight forward * require to effectively detect and analyze failures; context-specific; go beyond superficial * > need to throw the old cultural beliefs and stereotypes of success & failure

The blame game
Old cultural belief: admitting failure = take the blame —> no psychological safety
X a false dichotomy
√ a culture that makes it safe to admit failure can coexist with high standards for performance.
- why? the pic shows the causes of failures
It is hard to find real blameworthy actions * e.g. Inattention - caused by lack of effort vs manager’s arrangement
In many organizations, 2%-5% of failures detected by managers are truly blameworthy, but 70%-90% are treated as blameworthy.

Not all failures are created equal
1.Preventable failures in predictable operations mistakes in processes and routine operations * ususally caused by inattention, lack of ability * Solution: 1.Checklist * 2.Toyota production system - build continual learning from tiny failures
2. Unavoidable failures in complex systems
Often caused by uncertainty of work - combination of needs , people and problems never occurred before * e.g. emergency room in Hospital; start-up; complex system- nuclear power
Can be averted by risk management and safety practice, but small process failures inevitable.
Solution: rapidly identifying and correcting small failures, don’t let them line up.
3. Intelligent failures at the frontier is actually valuable cuz provide new knowledge that help with competition in the future * usually from experiment e.g. research on new product, new business, testing customer reactions… * the right kind of experimentation produces good failures quickly. * e.g. IDEO- launch new innovation strategy service - help client create new lines - failure-learn from the failure and hired MBAs and client’s managers- now 1/3 revenue * tolerance of unavoidable process failures in complex systems and intelligent failures.

Building a learning culture (in contrast with blame game)
Leader responsible and capable of creating the culture * 1. dev a clear understanding of what happened rather than who did it when things go wrong * need consistent report of failures, systematical analysis; proactive experiment * 2. send right msg of the nature of work e.g. remind people in research and development work ”fail faster in order to success faster”, but not in manufacturing work * routine, complex, frontier - ensure right approach to learn from failure

Learn from failures through 3 activities: * detecting failure
Prevalent problem: many methods to detect failures, but most staff/managers reluctant to report.
The human tendency to hope for the best and try to avoid failure at all costs gets in the way, and organizational hierarchies exacerbate it.
Solution: reduce stigma of failure
e.g. Eli Lilly = holding failure parties to nor intelligent failed experiment

2. Analyze failure
Dig in the root causes and discover the wisdom within the failure
But usually done too quickly or avoided * this is an emotionally unpleasant process; cognitive limitations- fundamental attribution error * managers not admired for thoughtful reflection, but decisiveness, efficiency and action
Solution: use interdisciplinary teams with diverse skills and perspectives
- NASA - go beyond first-order cause (e.g. foam), dig root causes further (e.g. hierarchy) * Promote experimentation
Pilot new product/service: often design optimal conditions instead of representative ones, so pilot doesn’t produce knowledge about what wont work
e.g. : Telco- DSL (bringing high- bandwidth information to homes and small businesses over ordinary copper telephone lines) - customer service disaster
The problem was that the pilot did not resemble real service conditions: It was staffed with unusually personable, expert service representatives and took place in a community of educated, tech-savvy customers.
However, it is a new tech, uncertain and complex, customers variable (edu; pcs); would have tested with limited support, unsophisticated customers, and old computers.