Smart Performance Charts (SPC)

Have you ever tried to explain control charts to a blue-collar worker or a CEO who has never seen one before? It’s not easy, is it?

What if it could be?

Over my 30 years in Quality Improvement, I’ve learned Japanese words like "gemba," "kaizen" and "Ishikawa." I’ve had to learn a slew of other jargon such as “Six Sigma" (complicated), but none more death-defying than "Statistical Process Control."

Statistical (ugh!) Process (boring) Control (hey, I hate control).

Words matter. Let’s face it, Shewhart did us no favors when his obsession with controlling manufactured quality spilled over into naming his creation “control charts.” In spite of their usability in endless applications, after almost 100 years few people know how to use them... let alone know how to use them effectively.

Dr. Donald Wheeler has been using the phrase “process behavior charts” to describe control charts. Again, both "process" and "behavior" are both boring words. Not very sticky. Lloyd Provost calls them “Shewhart Charts.” (Who the heck is Shewhart and why should I care? If they were called "Springsteen Charts" they might have a better chance of being remembered and used.)

In an age when everyone has a smart device tracking their performance (e.g., steps per day or heart rate), why are people so turned off by SPC? I’d like you to consider that the language of quality is off-putting. Jargon, in general, is off-putting.

Just as Mikel Harry launched a new phrase (Six Sigma) to replace TQM (Total Quality Management), which replaced quality circles and so on, it’s time to rename our methods and tools in a way that will make them more accessible, more desirable.

What if we changed the name of control charts?

Smart Performance Charts (SPC)

Smart...

First, let’s call them smart; who doesn’t want to be smart? These "smart charts" use statistics to help us detect anomalies in performance, much as a "smart watch" can detect atrial fibrillation. You don’t have to be a cardiologist to use a smart watch to detect heart problems, you just have to own and wear the watch.

Similarly, you don't have to be a statistician (or even a Lean Six Sigma Black Belt) to let a smart chart detect problems in your process. Modern SPC software such as QI Macros® makes it easy to create these charts instantly and update them just as quickly.

Let’s not pretend that everyone has to memorize statistical formulas just because we had to do this in order to create charts manually in the last century. Nonsense. In the last century, our great-grandparents had to learn to saddle and bridle a horse in order to travel; our grandparents had to learn to use a clutch and manual transmission in order to drive. Today, we can choose to ride a horse or drive a stick shift, but technology has made learning these things optional, and no longer a necessity.

There’s an app for that. Buy one. Start using it.

... Performance...

Next, let’s use the word “performance,” because everyone wants to improve their performance. Once you start keeping score, everyone will want to improve their performance. It’s human nature.

... Charts

And finally, it’s still a chart, just a Smart Performance Chart.

Unfortunately, most people are still using Excel line or bar charts. These charts are not that smart. They can’t tell you whether a process is stable and predictable, or unstable and unpredictable; Smart Performance Charts can. They can’t filter signal from noise; Smart Performance Charts can. They can't differentiate between special causes and common causes... but Smart Performance Charts can.

More importantly, Quality Improvement is about reducing and eliminating waste, and line and bar charts actually create waste. Line and bar charts can't tell you whether an apparently unusual point or series is within normal variation or something that's statistically unlikely. Because of this, fluctuations in line and bar charts often send businesses off on wild goose chases to determine why a key performance indicator (KPI) is up or down. Smart Performance Charts could prevent this waste of time and effort by showing that the process is stable and the fluctuation is within established limits. Companies around the world waste millions of hours every year chasing phantoms in their data.

They wouldn’t waste all this time if they used Smart Performance Charts instead.

SPC (Smart Performance Charts) 101: What do you need to know?

I've made the assertion above that most users today no longer need to memorize statistical formulas. We had to do this in order to create charts manually in the last century, but technology has made this no longer necessary. (Whenever I make this assertion, a large portion of the comments consist of people taking issue with that assertion... we'll see if this is the case here.) If we accept that assertion for the moment, what do most users need to know in order to use Smart Performance Charts?

What I think is more important than memorizing statistical formulas is people learning the patterns you're trying to achieve with Smart Performance Charts:

• What those charts are telling you, and more importantly,
• What you'd like them to look like after improvement.

Pattern 1: Reduce variation (tighten up the limits)

In Quality Improvement, reducing your process' variation is a key goal. If your process is unstable, reducing variation can bring that process into stability (and without a stable process, improvement is difficult... if not impossible). If your process is stable, reducing variation can lead to reductions is wasted time, wasted effort and wasted expense.

That's all well and good, but what does reduced variation look like in a Smart Performance Chart?

Your process' variation is represented by the distance between the UCL and the LCL. If your process change is effective, you should be able to see a distinct "before" and "after." (If your variation "after" the process change is implemented doesn't look any different than the "before," I have bad news for you: you didn't fix anything.)

Pattern 2: Move the Center Line down or up

Another key goal of Quality Improvement is improving outcomes in your process: increasing number of parts produced per day, decreasing number of defects per day, etc. Talking about that goal is one thing, but what does that look like in a Smart Performance Chart?

The answer lies right in the center of your chart. The Center Line represents an "average" or "median" value of the metric you're tracking: number of parts produced per day, number of defects per day, number of patient falls per week, etc. Moving the Center Line down (if you're tracking negative metrics like defects or patient falls) or up (if you're tracking positive metrics like number of error-free parts produced or patient satisfaction ratings) after a process change demonstrates an improvement. (Once again, if your Center Line "after" the process change is implemented doesn't look any different than the "before," I have the same bad news for you as before: you need to go back to square one and start again.)

Gone are the days of using gut-feel, common sense and trial-and-error to guide improvements, an approach that often causes more harm than good. Smart performance charts can help dramatically improve performance of any business process.

UCL/LCL – Upper and Lower Concern Limits

While upper and lower control limits are the common terminology for the statistically calculated limits, they don’t really control anything do they? Instead, what they do is raise concern if there is a point above or below these limits.

In a stable process, 99.97% of all points should fall within the UCL and LCL: a point outside of these limits should occur just three times out of a thousand. If there’s one point outside these limits in a data set of 20-50 data points, it should raise concern and be investigated as a possible special cause. Let’s call these limits what they are: concern limits.

The one- and two-sigma lines within the UCL/LCL are also areas of concern. Two out of three points above or below the two-sigma lines are a cause for concern. (Imagine taking three photos in quick succession of the car in front of you: one photo shows the car very close to the left side of the lane and another photo shows the car very close to the right side of the lane. While the car has technically stayed within its lane the whole time, the car's inconsistent path should raise some concern...) Similarly, four out of five points above or below the one-sigma lines are a cause for concern as well.

I’d suggest introducing UCL/LCL to new students as Upper and Lower Concern Limits.

"Six Sigma"

Mikel Harry did us no favors when he named Motorola’s approach to quality “Six Sigma.” Sigma is statistical jargon, and jargon makes people balk. With Six Sigma waning in popularity, maybe it’s time to coin a new phrase.

In keeping with current business lingo, I’ve started using the phrase “Agile Process Innovation” to describe what we do. "Agile" feels fast and athletic to me. "Process" (boring) plus "Innovation" (exciting) could be more approachable and less scary to the uninitiated. To learn more about Agile Process Innovation, download my free ebook HERE.

I also considered using the term “Agile Lean Six Sigma,” but it carries the old (Sigma) with the new (Agile). Of course, people who already know about Lean Six Sigma will be intrigued by this name. People who don’t may still be put off. "Agile Process Innovation" is currently unoccupied space in fresh minds.

Conclusion

Gone are the days when we could force people to spend weeks in training to learn complex methods, tools and jargon to engage in performance improvement. In the interest of spreading the gospel of quality, let’s consider renaming our methods and tools to make them more desirable and sticky.

We all stand on the shoulders of giants, but the Quality Improvement giants are gone now. It’s time to stand on our own two feet.

For more information on how to make ideas sticky, consider this resource:

Made to Stick: Why Some Ideas Survive and Others Die, by Chip and Dan Heath

Jay Arthur teaches business people how to Turn Data Into Dollars® using QI Macros for Excel. He is the author of Agile Process Innovation, Lean Six Sigma Demystified, and Lean Six Sigma for Hospitals. He can be reached at jay@qimacros.com.