Thursday, June 14, 2012

Information Costs and Six Sigma

Probably one of three most important concepts I learned at the Simon School was information cost. (The other two were: Buy and hold a diversified portfolio, and sunk costs are irrelevant. There -- you just got the benefit of an MBA education. Go forth and prosper!) Information cost is the costs that occur because the person making the decision is too far away from the process to understand exactly what is happening. In many organizations, not only do the misplace decision makers not know the impact of the decision, but even worse, they don't know what they don't know.

So, when applying Six Sigma to a process, one critical place where the whole thing can go south in a hurry is having someone make decisions about sensitive systems when the person doesn't truly understand the system.

Here's an example from my past. I worked for a company that made, among other things, e-learning. The courses were divided into lessons, the lessons into topics. The content was complex. If it were easy, people would have figured it out on their own. They wouldn't need us to teach them how to do it.

Typically, we worked on a lesson basis. The process started with a planning meeting, and then was written, edited, and animated. This mean that there was some time waiting in for the content to get to you, especially if the previous lesson was smaller, or for some reason you finished the work early. Management watched utilization very closely, and decided that handing off lesson-by-lesson resulted in too much waiting, and we should hand off topic-by-topic.

The problem with that is that the entire process had been built with the assumption of lesson-level hand offs. We were accustomed to moving content around in lessons, adding or deleting topics, and so forth when the lesson was edited. When we were forced to edit topic by topic, once the topic went by your part of the process, that was it. If you got to a later topic and made an edit that affected (or need to affect) something in an earlier topic, well that was just too bad. As the content reviewer, I started having to pull back early topics. This resulted in re-work all the way down the line. It was a nightmare. At first management didn't want to hear anything about going back to a lesson-level hand off system. They stubbornly clung to their decision. But eventually their own figures  on cycle time convinced them to reverse themselves. What a happy announcement that was!

The take-away is, I think, that we have to approach process improvement with humility and collegiality. We need to watch the impact of our changes, and be ready to tweak as necessary. And we need to think about the unchanged systems affected by this process. It is only by attention to these that process changes will be successful, and not include some unexpected and unhappy results!


Post a Comment