What is failure? When things don’t go according to plan or expectations, ending up with unexpected and/or undesired outcomes (which we can argue could have been avoidable, or not). The key is ‘undesired‘ – because if they were desired and not planned or expected, that would still be great! But, as we will see, failure is a terrific way to learn. Maybe we could measure learning as Return on Failure: ROF.
We’ve all heard the phrase “fail often, fail cheap, fail fast.” So, can we do a better job of learning from failure? We’re not built to do this easily, either by learning from others’ failures or our own. There are many ways to learn from failure, so what I’m suggesting is just one way.
One way we could start learning from failure is through a simple 3-step process (bear in mind, simple ≠ easy!):
- Identification of the Failure(s)
- Analysis of the Failure(s)
- Iterative Experimenting & Prototyping based on the learnings from the failures
So, and check my ‘math’, ROF is the sum of Failure Identification + Failure Analysis applied over (and over…) Iterative Experimenting & Prototyping. That’s the framework (for now).
ROF = (FI + FA)/IEP…
Failure Identification (FI) is proactively identifying what went wrong, what failed. Systems and processes can help capture this information for sharing with those who need to know now and in the future. Feedback loops with employees, customers, and suppliers are also important (and who else?). Most companies are complex entities which make getting and sharing information difficult. Also, most cultures don’t tolerate failure too well so we learn to play the blame game. And of course, there are a lot of other reasons we’ll get into in further posts.
Failure Analysis (FA) is not playing the blame game but discovering the Why. When a plane crashes, the NTSB goes over every inch of the site. They don’t blame; they use a formal, objective process to discuss, analyze and learn. Try a model like this. Be objective, don’t personalize or blame (not as easy as it sounds). Organizations also succumb to confirmational bias; we become inured, not realizing we’ve fallen into that trap. The “blame game” makes doing the necessary forensic work challenging because it can be hard to trust our colleagues.
Iterative Experimenting & Prototyping (IEP) involves creating a well designed experiment so we can limit and test the variable (ideas) and prototype. Test where we think we could fail, try what does and doesn’t work. The more we experiment, the more we learn, the greater the chances of success. Do small, inexpensive experiments and prototypes (they don’t have to be grand). Do virtual and thought experiments. There are many ways to experiment and prototype today that are not expensive or lengthy so try it. Why don’t we? How many organizations are structured for experimentation? Not many (remember the scientific method? Bet not). And culturally, we don’t incent, reward, recognize our people to experiment – we incent being right, not trying to be right!
What do you think? Does this make sense? Are you trying to learn from failure in your organization?
Stefan Lindegaard, one of Open Innovation‘s top Gurus, has called this area of learning from failure smartfailing (see comments to this) and has created a place for discussing Smartfailing, which I encourage you to check out and share.