Understanding why skunkworks projects fail

- Baseball's got nothing on entrepreneurship when it comes to failure. If you're batting .300, striking out more than two out of three times, you're a star. With disruptive innovation, you're going to fail even more. That can be demoralizing. One way to make failure easier and less painful is to define failure. There are a two kinds of failure. One is scientific failure. You set a hypothesis of what you think will happen, try it, and note the result. This kind of failure is valuable because it allows you to rule out ideas that don't work so you and your colleagues don't waste too much time on it. The other kind is careless failure. This failure is due to poor execution on the part of the innovator. Sure, you've used sticky notes and colored markers and developed some big ideas, but then you fall short. You don't clearly define your hypothesis. You don't keep good records. You start the project but don't have the resources to complete as planned. You don't get support and help from your peers when needed. If you're in the skunk works phase, that time when you have a small green light to begin experimenting, focus and discipline will be critical. The prior phase may have been all about ideation, creativity, and pie in the sky brainstorming, but now you need to come all the way back down to earth and scope something small and measurable. Many innovators have trouble toggling between these two states. If you're unable to translate your creativity into modest tests, you doom your project and risk losing credibility as someone who can make things happen. Many organizations encourage skunk works activities. They provide all employees with time and even some education on ideation and creativity in hopes of generating exciting new ideas. If you want to avoid unacceptable failure before transitioning from idea selection to idea testing, make sure you scope your project. What is the hypothesis you're testing and what would success look like? What will you learn that will justify the next investment? You might be testing whether the technology works or whether a particular customer segment will react positively, or whether a new process will improve results. Whatever you're testing, scope it as small as you can, but no smaller. In other words, the test should be fast and small but big enough to be seen as significant by your colleagues. It can be tempting to allow your scope to increase bit by bit, but scope creep can really slow you down and muddy the water too. Additionally, you need to be honest and thoughtful about what resources you'll need. Of course you want to be lean, but don't underestimate the time, money, and support necessary to get a conclusive result. Undersell, overdeliver. Set a date by which you will know your results. Educate any stakeholders of what would denote success and failure, and when that day comes, proudly share your learning. It's okay to even identify additional tests you need to do or resources you need to add based on learning. Remember, it's not failure if it moves you forward. Remember, the biggest failure in any skunk works effort comes from not starting at all. Don't let that by you.