Tuesday, February 28, 2012

The Improvement Process

Credit: Mathews & Co.: "How're We Doing?" TM 

I like this chart from Mathews & Co.'s "How're We Doing?"website, and I wanted to spend a few minutes describing what I like about it and what I might do differently. This is my unsolicited critique of a particular attempt to define the improvement process.

Let me begin by saying that I do not currently have any dealings with Mathews & Co. or "How're We Doing?" but we seem to have a common interest in Continuous Process Improvement (CPI) and in what I have called Data-Fueled Decision-Making (D-FD-M).

According to my reading of the chart, the folks at Mathews & Co. define process improvement as having the nine following traits. My verbiage reflects my understanding of and degree of agreement with the original text.

  • Goal oriented. It's very important to have clarity on the significance and urgency of a desired outcome.
  • A team sport. Not just a team, but a balanced team--different perspectives and skills working in harmony (Belbin). 
  • Not easy. Change is never easy and organizational change is especially hard. To improve a process is to change culture, i.e., "the way we do things around here." This requires top-down support (Kotter).
  • Data-fueled. Not all processes are data-intensive, but even simple processes can be better  understood with data (Deming). 
  • Logical. Effects have causes. If the process results are not consistent or not as good as desired, examine the causes of variation and error.
  • Creative. You cannot expect to solve a problem using the same thinking that got you in trouble in the first place (Einstein). 
  • Competitive. 
  • Consuming resources and carrying risk. 
  • Cyclic. Process improvement is not linear. Rather it is cyclic and never-ending.  

As much as I like the chart and would keep what I highlighted above, I would add a couple things and do a couple things differently.


  • There are some aspects of CPI that are so critical to discuss that their absence from the chart is disappointing. Perhaps these missing topics are addressed in the details, but again, I would give project selection, portfolio management, assessment frameworks, and dynamic system modeling their own bullets. 
  • CPI is best thought of as a campaign conducted enterprise-wide. General Electric does not exist for the purpose of improving itself, but you had better believe that in GE's culture, CPI is tied to profitability, competitiveness, and innovation. Whether the score is kept in stock prices or on an assessment framework. all stakeholders have an interest in changes being made to any part of the enterprise. 
  • Process improvement projects should be selected based on expected enterprise-wide return on the time, talent, and treasure (T3) invested. Every process can be improved, but it makes sense to identify the constraint (Goldratt) and launch coordinated projects that reduce that constraint.


  • In the illustration in the upper left of the chart, I suppose there is a reason why the person on the highest bar is facing the other climbers. The implication is that he has arrived at the top. He may even be looking back on where he has come from to get to the top. However, the image works against the notion of continuous process improvement. I would prefer to see the person continuing to climb bars that fade to infinity. Onward and upward! 
  • I would bring the arrow head back around to the goal, to indicate that CPI is a journey, not a destination, and CPI is a loop, not a line. 
  • I would de-emphasize the linear arrangement of the bullets. The steps may be done in order, but  since the process is non-linear and cyclic, it does not really matter.

Again, I am interested in CPI and I like this depiction of it. What are your thoughts about CPI? Using this chart from "How're We Doing?"as a point of departure, what would you sustain, start, or stop? Comments welcomed!