Scott Berkun did a great job summarizing a talk Ed Catmull gave at a recent Economist conference. In Scott’s post were several transcriptions, one of which stopped me in my tracks:
The notion that you’re trying to control the process and prevent error screws things up. We all know the saying it’s better to ask for forgiveness than permission. And everyone knows that, but I think there is a corollary: if everyone is trying to prevent error, it screws things up. It’s better to fix problems than to prevent them.
W. Edwards Deming must be spinning in his grave. Catmull is saying precisely the opposite of what Deming, Juran and others began preaching in the years after World War II and which became ubiquitous in the 1980s and 1990s. The quality gurus taught us that detecting errors in manufactured products after they had occurred was expensive and wasteful. It was far better to identify problems as they occurred, “pulling the andon cord” if necessary to isolate and solve problems so that superior quality was baked into the product throughout the process, not created in a later inspection step.
This was a brilliant insight and helped revolutionize manufacturing, enabling just-in-time inventory, lean production and Six Sigma. But something else happened. Six-Sigma-like quailty processes were extended to processes like marketing, sales, product development, etc., processes in which variance was not necessarily a bug (in some cases it was a very valuable feature).
Catmull’s business is the polar opposite of the high-precision world of mass manufacturing. As he says in the interview, during the beginning of a film development process, “everything is broken.” You can’t measure conformity with specs because there are no specs. And variability is essential to the process. Only by putting lots and lots of wild ideas into the top of the development funnel do you have a chance to come out the other end with a work of art.
So his words are very helpful to think of when we are working with our business. Is this a conformity-based process? Is error bad? Or is it an artistic process, in which error is inevitable and, in fact, essential? When is it the situation of, “if everyone is trying to prevent error, it screws things up. It’s better to fix problems than to prevent them”?
The entire interview is here (the section I quoted is near the very end):