3 Steps to Building Your Own Innovation Machine (Part 3)

Here comes another question – how can we better learn from the experience we get? Eric Ries, already mentioned above, uses an efficient way to tackle problems. I am talking about root-cause analysis or “five whys.” Imagine that the problem you’ve faced has the same structure as a Russian doll. The “root cause” of it is hidden inside, and you have to remove several layers to get to it. Just as you take one doll out of another, you ask a question “Why did this happen?” five times. Each response takes you one layer deeper to the problem cause. The technique is quite easy, but when practiced regularly, it gives you a lot of great insights about what needs adjustment in your company. One of such insights is that there is always a process/human issue behind every technical one. For example, imagine that you increase the investment in advertising, but the return from it is not proportional. Why did that happen? It seems that the quality of leads dropped. Why did that happen? Because we didn’t have a quick feedback loop between the money put into advertising and the output from advertising that we get. Why didn’t we have that feedback loop? Because we didn’t know how to properly score leads. Why didn’t we properly score leads? Because no one did statistical data analysis. Why didn’t we mine the data? Because the process didn’t allocate time for someone to periodically mine the data. The next step of implementing root-cause analysis is to make a proportional investment to correct each level of the problem. It helps you to avoid both ignoring and over-reacting to a minor problem. In the case above, the decisions could be to allocate some time for data mining, to score the leads, to feed that data back to analytics, and to adjust advertising campaigns based on that. You address the problem on all levels with an incremental improvement. Every time you face a problem, you make an effort to improve the company at multiple levels with small steps. If the problem is more complex than you thought, it will keep occurring, and every time it reoccurs, you will make an incremental improvement, until it is finally solved. This way, you invest your time and money only into the part of your business that needs it the most. “Lean startups” are lean. You can see that this method combines learning and doing, continuously changing your company as you learn. This fits perfectly in the frame of continuous learning. To close, I also want to mention the importance of your own unique vision for the product. I don’t mean to suggest that the shortest way to success is to simply follow every customer’s request. Those requests will often pull the company in different directions, and you don’t want to be doing Brownian motion. This is where the art mixes with the science and produces brilliant results. What about your professional experience? How do you normally deal with failures in your business?