Improvement Insights Blog

In the early 1990s, I took a control chart class that lasted five days. We learned decision trees and how to calculate control chart formulas with a calculator. I knew that the phone company would never tolerate five-day trainings for 70,000 employees and I knew that few of our employees would tolerate calculating these formulas manually. The only way to remove this barrier was to automate it. Here’s how I did it.

Let’s stop teaching people things they don’t need to know. It’s waste. When I automated everything, it shortened the training and increased retention and results. You can too.

“Hi, I’m Jay Arthur.

“In the early 90s, I took about a week-long control chart course where we had to spend time studying decision trees and trying to figure out how to learn things, and then we had to use calculators to calculate the formulas and all that kind of stuff… Well, that was our experience back in “last century quality.”

“Recently I presented this bit of information where anything that’s repetitive and cognitive (or thinking or mental), is all going to be taken over by software algorithms. When I left the phone company in 1995, in 96 and 97 I needed some tools to start to do some of this work with clients when I was doing consulting, and so I started building the basic parts of the QI Macros [software]. First I created control charts, and so I automated all the formulas so I didn’t need my little calculator any more.

“Then we had sort of a “decision tree-like” thing that you had to navigate to get into doing a control chart. Then in 2006 I had this epiphany: I could automate that, so in 2006 the Control Chart Wizard was born. QI Macros is the only tool that allows you to select data first, and once you selected it, then the Control Chart Wizard can look at it and choose the right chart for you. So then I didn’t have to teach people anything about that. The funny thing about decision trees is that’s not a natural way of thinking for humans, and so it was very hard to teach and they often couldn’t remember it. They’d [ask,] “Jay, what kind of chart do I use with my data?” Well, select the data, click the Control Chart Wizard and off you go.

“A few years later, I found the same sort of problem with statistics; nobody knew what to do with this data. Well, I created a Stat Wizard and the Stat Wizard will go through the decision trees based on what kind of data you have. What used to take me a while to try and teach people has been collapsed into “Select data, click a button, get a chart.”

“This is how we start to hack Lean and Six Sigma training. As we automate the things that are repetitive, that people don’t need to know. Then I also did a Chart Wizard, and the Chart Wizard will look at your data and [think,] “Well, what would I do with that data?” It could be anything from a box-and-whisker to a control chart to a Pareto chart. You click that button, it’ll look at your data to say, “Oh, I think we could do these charts with that data.” That simplified a lot of things, because I used to get a lot of tech support calls, and they’d [ask,] “Jay, what do I do with my data?” Well, select the data, click the Chart Wizard.

“Then, along the same time, I had learned or figured out in the mid-2000s that I was using pivot tables, and pivot tables were the key to all multi-million dollar improvement projects, because the invisible low-hanging fruit is in raw data. So what I had to do is teach people how to use pivot tables. Well, most people didn’t know how to use pivot tables. So I created a Pivot Table Wizard which would take your data and put it in the right format so that you could actually start to draw control charts and things like that with it.

“Now, it was about ten years later where I had the epiphany: “Jay, you dummy, what you’re doing is an algorithm.” So I created the Data Mining Wizard which will take your data – your raw data – and pivot it and control chart it and Pareto chart it and drill down and add in fishbones; it’ll build an entire improvement project for you.

“So it used to take me three days to do train, then took me two days, and now I’m down to one day, and Six Sigma is about four hours of it because I let the software do the heavy lifting. Now, what does that do? Well, that allows people to very quickly get up and going doing Lean and Six Sigma, and they’re all asking that question, “Will it work for me? Will it work here?” If I can very quickly take them through their data and turn that into an improvement story so they get a sense of what is possible, then they just take off and run with it.

“This is how we start to hack Lean and Six Sigma training, right? We embed as much as possible into the software so you don’t have to teach it in the classroom, and if you’re teaching formulas and decision trees in a classroom or any of these other kinds of crazy things, you’re wasting everyone’s time. Let the software do that for you.

“That’s my Improvement Insight for this week. Let’s go out and improve something. “