Short talk arguing that ‘failure’ should be reframed as experiment outcomes. The speaker presents four possible experiment outcomes (confirm hypothesis; disconfirm but reveal new insight; disconfirm and kill the idea; inconclusive due to insufficient data) and claims only the inconclusive outcome is true failure. Framing work as experiments lets organizations contain risk, avoid glorifying failure, and make innovation more cost-effective.
Chapters
In the world of innovation, there is this belief that if you want to make an omelet, you have to break a few eggs. Or in other words, failure is inevitable. Or as Adam Savage from TV's Mythbuster says, Failure is always an option. And yes, failure is always an option. However, in many organizations, failure is being glorified, which I think is a mistake. And also, if we redefine what failure means, maybe it does not become so acceptable. See, in the world of innovation, one of the best things we can do is view everything through the lens of an experiment. Because when we run experiments, we are able to contain the risk. And then we can use a build it, try it, fix it, model, build something, try it out, see if it works, adapt it, and then scale it. So if we're doing experiments, there's a number of different outcomes that we can have. The first outcome when we run an experiment is that we have this hypothesis we want to test. And in running the experiment, we confirm the hypothesis. Fantastic. We know to invest more money because we're on the right path. Another outcome, we have a hypothesis, we test the hypothesis. That one didn't prove to be so true. However, we discovered something else in the process. We have a new hypothesis, so we start to build experiments around that. The third option, we have a hypothesis. We tested a hypothesis, guess what? It was wrong. We killed the experiment. The last option, we have an experiment, we test the hypothesis, and our experiment does not support or refute the hypothesis. We didn't gather enough data. Now, of those four, it is only that last one that is truly failure. The experiment only fails when we do not get enough information to confirm or refute our hypothesis. The third one, where we had a hypothesis and it was disproven, is actually an incredible success. Killing off innovations is as important as supporting new innovations. So when we redefine failure and we think about failure through the lens of experimentation, it gives us a fundamentally different perspective and allows us not to glorify failure, but to contain failure and to do risk mitigation strategies that will enable innovation to emerge in a more cost effective way.
