Reflecting on Lean
I found this half-written note while cleaning out my Evernote account earlier this week. The note was from July 2013. My team had been operating within the lean methodology for about one year and I was reflected on some of the challenges. Interesting to look back on that time.
Below is the unedited *draft* version of my note.
Let me start off by saying that I’m a firm believer in the lean methodology.
I believe in it. I practice it. I’ve seen it work.
I’ve also seen it not work… And it has nothing to do with the talent of the team.
The problem arises when lean is used as a shortcut.
Repeat after me… Lean does not mean throwing shit against the wall to see what sticks. Lean is not a license to bypass a strong product vision or an excuse for poor quality.
If you don’t take the time to develop a strong pov in your product, you’ll end up building a muddled confusing experience for the user and you will fail. And at that point all you’ve learned is that customers don’t like muddle.
Done correctly, Lean challenges you to develop a strong pov… Your hypothesis. You cannot skip that step and expect clean, measurable results.
You also can’t let lean result in low quality. This is actually an easy trap to fall into with an MVP. By definition an MVP doesn’t have to be perfect. It just has to be good enough to test the hypothesis.
However, there are two problems.
First, this can become a slippery slope where buggy features with design flaws become accepted in the team culture. I’ve seen it happen. Pretty soon the icebox is full of low priority bugs and design debt.
The other danger is that of shifting priorities. Unless executive management is bought into the idea of iterative learning, teams will get pressure to move on to the next high priority feature without iterating on the one they just built. And if you don’t keep iterating on the feature after MVP launch, then you don’t really have an MVP, but just a crappy product.
Quality should not be a casualty of fast iterative learning. The key is to move quickly with quality in the DNA by limiting the scope of the feature.
Lean sounds like common sense, but it is hard to do correctly.