Why I dislike the phrase “Fail Fast, Fail Often”
How many times have we all heard the modern words of wisdom that tell us to embrace failure? Well, I simply cannot afford to fail. When starting a new business as the sole earner in a family of five, failure is not something any of us will be celebrating. I’d much rather succeed. Failure sucks.
Before anyone jumps down my throat, I should disclose that I used to co-lead the Agile Guild for a large global consulting firm, so it is not as if I need to be convinced of the benefits of an iterative approach to doing work. I fully believe that a linear waterfall approach in the face of uncertainty is likely to be a recipe for disaster. It’s just that executives get paid lots of money to not fail. Shareholders in organisations tend to punish failure… and so they should. Imagine two competitors where one fails fast and often while the other succeeds fast and often. I know where I would want my money invested.
I of course understand what the phrase is trying to convey. And, I appreciate the sentiment. It is far better to make a mistake early through a test and learn approach and adjust as a result. Yes… fully agree. I also understand that “fail fast fail often” is trying to de-stigmatise failure in a highly uncertain context to fully encourage experimentation. However, lots of organisations are kidding themselves in that they experiment, fail, learn nothing, repeat the same errors and fail again. An iterative approach to work where mistakes are seen as an early warning mechanism is only effective if mistakes translate to learnings which help avoid future mistakes. If something breaks in that chain, you can get stuck in a world of fail, fast fail often… and perhaps fail permanently.
I know it may sound like mere semantics, but I tell my clients that we want an environment of “learn fast, learn often”. It is ok to get things wrong but only if you learn from those mistakes and adjust as a result. What I find incredibly frustrating is repeating the same mistakes over and over again.
Unfortunately, many organisations seem to be stuck in a mode of learn slow, fail often. One organisation I know trumpeted a commitment to building MVPs, but looking under the hood showed that an MVP took 18 months and cost approximately $3M to fund. When I spoke about prioritising the collection of some data to provide just enough information to determine whether to move forward or not, and that the financial commitment to do this could be measured in the tens of thousands, they simply didn’t believe me.
I honestly believe that the words are important. “Fail fast, fail often” tends to make more conservative executives nervous. In their realm of experience where “fast” might look an order of magnitude different to what I have in mind, my sense is that the only word that really sticks is “fail”.
The phrase also emphasises the negative. My preferred “learn fast, learn often” takes the spirit of iterative work and asks executives to think through how they translate iterations and insights into learnings. Doing this requires more than sprints and Kanban and retros – what I sometimes refer to as “small a” agile. The only way to institutionalise learning is to think through what needs to happen around funding and resource allocation (dollars and people), governance, planning and budgeting, and even your talent environment. If failing fast and often when it comes to lots of small but highly ambiguous and unknown project environments stalls career progression, while delivering on a large and safe known capital project is rewarded, what message does that send around organisational priorities? Where will your most talented colleagues align themselves?
“Learn fast, learn often” requires an organisation to think about all the invisible enablers that most people simply take for granted. The learning box is only ticked when you can demonstrate that something has changed. The failing box is ticked when something hasn’t worked, and it is therefore a much lower bar.
Creating a learning organisation is suddenly in vogue. It is rightly seen as a critical part of being an adaptive organisation and the key to remaining responsive in the face of constant change. Working in iterations is simply not enough. Changing our language from that of failure to that of learning may be seen by some as the flip side of the same coin, but in my experience, what appears to be a trivial language change can ask powerful questions of an organisation that go as much to mindset as process.
I don’t want to fail. The only way my business will succeed is to make lots of mistakes, make them quickly and cheaply, learn from them, and avoid making those mistakes in the future. Likewise, when it comes to finding new colleagues for Vibrance, I’ll be looking for a track record of learn fast and often. Repeated failure won’t get a look in.