The Lean Startup - Eric Ries [110]
Like many entrepreneurs, I was caught between constant evangelizing for my ideas and constantly entertaining suggestions for ways they could be improved. My employees faced the same incentive I had exploited years before: the more radical the suggestion is, the more likely it is that the compromise will move in the direction they desire. I heard it all: suggestions that we go back to waterfall development, use more quality assurance (QA), use less QA, have more or less customer involvement, use more vision and less data, or interpret data in a more statistically rigorous way.
It took a constant effort to consider these suggestions seriously. However, responding dogmatically is unhelpful. Compromising by automatically splitting the difference doesn’t work either.
I’ve found that every suggestion should be subjected to the same rigorous scientific inquiry that led to the creation of the Lean Startup in the first place. Can we use the theory to predict the results of the proposed change? Can we incubate the change in a small team and see what happens? Can we measure its impact? Whenever they could be implemented, these approaches have allowed me to increase my own learning and, more important, the productivity of the companies I have worked with. Many of the Lean Startup techniques that we pioneered at IMVU are not my original contributions. Rather, they were conceived, incubated, and executed by employees who brought their own creativity and talent to the task.
Above all, I faced this common question: How do we know that “your way” of building a company will work? What other companies are using it? Who has become rich and famous as a result? These questions are sensible. The titans of our industry are all working in a slower, more linear way. Why are we doing something different?
It is these questions that require the use of theory to answer. Those who look to adopt the Lean Startup as a defined set of steps or tactics will not succeed. I had to learn this the hard way. In a startup situation, things constantly go wrong. When that happens, we face the age-old dilemma summarized by Deming: How do we know that the problem is due to a special cause versus a systemic cause? If we’re in the middle of adopting a new way of working, the temptation will always be to blame the new system for the problems that arise. Sometimes that tendency is correct, sometimes not. Learning to tell the difference requires theory. You have to be able to predict the outcome of the changes you make to tell if the problems that result are really problems.
For example, changing the definition of productivity for a team from functional excellence—excellence in marketing, sales, or product development—to validated learning will cause problems. As was indicated earlier, functional specialists are accustomed to measuring their efficiency by looking at the proportion of time they are busy doing their work. A programmer expects to be coding all day long, for example. That is why many traditional work environments frustrate these experts: the constant interruption of meetings, cross-functional handoffs, and explanations for endless numbers of bosses all act as a drag on efficiency. However, the individual efficiency of these specialists is not the goal in a Lean Startup. Instead, we want to force teams to work cross-functionally to achieve validated learning. Many of the techniques for doing this—actionable metrics, continuous deployment, and the overall Build-Measure-Learn feedback loop—necessarily cause teams to suboptimize for their individual functions. It does not matter how fast we can build. It does not matter how fast we can measure. What matters is how fast we can get through the entire loop.
In my years teaching this system, I have noticed this pattern every time: switching to validated learning feels worse before it feels better. That’s the case because the problems caused by the old system tend