thomas

In aviation safety, we use James Reason’s “Swiss cheese” model to describe the successive layers of organizational defenses against human errors (which, themselves, form the overwhelming majority of errors). 1/

Aviator, diver, Etherean, nexialist, PhD in aviation ∩ climate science | Host of /diving /ethfinance /eth-staker /geopolitics /science /singapore /thomas

That’s why you often hear that aviation accidents have multiple causal factors; it takes a combination of failures to cause a crash. 3/

It is by multiplying all these successive low probabilities of “successful failure” that we get to the current >6 sigma (in six sigma parlance) level of aviation safety, or roughly 0.03 accident for every one million flight sectors. 4/

Now, constantly reading reports that glorify entrepreneurial success stories has made me think deeper about survivorship bias, and how the media (especially social media) magnifies success and either hides or minimizes errors. 5/

Elon can get away with a cheeky tweet about Starship’s “rapid unscheduled assembly”, laugh it off, and move on. More generally, successful founders attract strong audiences online, but rarely do we hear honest accounts of the many more founders who failed. 6/

By over-indexing on successes, we are not taking the opportunity to learn from mistakes.

Whereas in a safety-critical industry like aviation, nuclear, or pharma, we make sure to investigate and analyze every near-miss, incident, or accident, not to apportion blame but so we can learn from them and avoid reoccurrence (a “just culture” approach). 7/

That’s why I’m proposing to invert and repurpose James Reason’s model. Instead of seeing its layers as artificial defenses against errors, I think of them as natural obstacles to success. What are the typical factors trips up entrepreneurs and founders? Personal competence? Access to capital? Market timing? Poor PMF? Lack of a support network? 8/

If you flip the Swiss Cheese model on its head like I do, you can start thinking about how “the stars [or in this case, the holes] must align” for someone to be successful.

I think it can not only help establish a standard taxonomy of failure modes (“why do entrepreneurs fail”), but also help put into perspective how failure is actually the norm and success the exception, despite its oversampling on social media. 9/ 🎬

The idea is that each layer of the model is meant to catch a certain type of error, such that it is extremely unlikely that a failure will propagate through the weaknesses in every successive layer (the “holes”) and lead to a catastrophic outcome. 2/