Ran into this wonderful article on learning from failure by HBS professor Amy Edmondson:

Strategies for Learning from Failure

For many of us, failure often has a negative association. Mostly because from a very early age, failure is so tightly coupled in our minds with fault – admitting failure also means taking the blame.

And yet, in the business world, success almost always requires some level of risk taking – a situation in which failure is an inherently plausible option.

How can we tell “good” failure from “bad” failure,  so we can encourage the former and discourage the latter?

Amy suggests a taxonomy of failures ranging from the blameworthy to the praiseworthy:

1. Deviance – an individual chooses to violate a prescribed process or practice
2. Inattention – an individual inadvertently deviates from specification
3. Lack of ability – an individual doesn’t have the skills, conditions or training to execute a job
4. Process inadequacy – a competent individual adheres to a prescribed but faulty or incomplete process
5. Task challenge – an individual faces a task too difficult to be executed reliably every time
6. Process complexity – a process composed of many elements breaks down when it encounters novel interactions
7. Uncertainty – a lack of clarity about future events causes people to take seemingly reasonable actions that produces undesired results
8. Hypothesis testing – an experiment conducted to prove that an idea or design will succeed fails
9. Exploratory testing –  an experiment conducted to expand knowledge and investigate a possibility leads to an undesired result

More broadly, failure requires a different reaction in three major domains:

1. Blameworthy: preventable failures in predictable operations  (#1-5)
2. Neutral: unavoidable failures in complex systems (#6-7)
3. Praiseworthy: intelligent failures at the frontier (#8-9)