Earlier this week I had a conversation with Benjamin Mitchell about an entrepreneur he’d done some work with. Benjamin had been suggesting a lean start-up approach to the business in which the entrepreneur worked to validate his business idea early and gradually build the product one validated step at a time.
Things didn’t go well, the understanding I came to from Benjamin’s story – although his might be different – was that the entrepreneur was very attached his business idea. It was pretty well formed and he could imagine it working. In popular culture entrepreneurs are often seen as iconoclasts who battle doubters, naysayers and banks to bring their ideas to life against the odds. Benjamin was just another doubter.
Now I tend to agree with Benjamin’s approach because it is very difficult to tell if a new business idea, indeed any project, will work until you try to build it. But building it is expensive so I advocate a “Try, fail fast, fail cheap, move on to the next thing” approach – I don’t think this is that different to lean start-up thinking but as I’ve not read the lean start-up book I don’t really know.
I don’t just advocate this approach for new businesses, I advise it for established businesses embarking on a project. My logic is: you can’t be sure a product will work in the market till you try it, nor can you plan a project until you know how fast (velocity) the team will work. In fact, until the team trys to work together and build something you don’t have any data on which to project plans. Making estimates without data is little better than guessing.
Thus, I advise a Mao like approach to corporate projects, “let a thousand flowers bloom” – start multiple small projects, small teams and use portfolio review to remove those that don’t seem promising and reallocate resources to those which do seem promising – or launch more experiments.
Unfortunately, like the entrepreneur, people get very attached to the idea of a project so it gets momentum. Big corporations respond by only starting projects in which they have a high degree of confidence. That means they do a lot of pre-project work – planning, designing, etc., this in turn makes it difficult and expensive to start projects, which in turn means that any project that does get started already has momentum. And it means that once underway the project finds it difficult to cope with change.
It seems to me that successful projects in corporations are often the result of one person who wants the project to happen and does what is necessary, not unlike entrepreneurs. In both cases a strong willed, capable, person makes something happen. That person needs to be motivated, they need hope, they need ambition and hope. Such people are going to kind who ignore naysayers and doubters.
The Agile approach – try, gather data, evaluate, or my “fail fast, fail cheap, retry” approach are not going to go down well with the kind of pig-headed, obstinate, passionate, do-anything-it-takes, approach which has made many successful projects a success. Quite the opposite.
Interestingly Benjamin had another example of a company which does take the “fail fast, fail cheap, try again” approach and was staffed by passionate, hopeful, people. From what he told me these people were passionate about the process as much as they were passionate about the products they tried. They could accept regular failure of good ideas because they saw the process working, and perhaps more importantly, could transfer their hope to the next idea. There was always a next idea.
What I am saying is: Don’t rely on hope to get your projects done, use data – be empirical not emptional.
Finally, I should say, that while I talked much of this over with Benjamin he would be the first to point out unvalidated steps in my thinking, and ungrounded assumptions I’ve made. So maybe this entire entry is really a hypotheses. The hypothesis is:
“The trial and error approach of Agile is unlikely to please those who get passionate about an idea and want to move heaven and earth to make it happen.”
I’ll let you, dear reader, decide for yourself on that hypotheses. If you have any evidence – to support or disprove – my hypothesis please add a comment below.