Executives have told me process improvement methodologies such as Lean, Six Sigma, and TOC are too complicated for their own daily use. They say these methodologies require too much time for mastery and aren’t fast enough for their reality.
I’ll grant some methodologies are built for thoroughness over day-to-day usefulness. I’ll also admit folks in process improvement can be so proud of our shiny tools we feel compelled to tell everyone about each one in detail.
And executives have also told me that hell is having someone telling them more than they want to know, and taking an eternity, or three minutes, to do it.
So I challenged myself with finding the “simplicity on the other side of complexity” for a process improvement presentation at a retreat hosted by the Institute for the Advancement of Behavioral Healthcare.
To make my goal clear I called my talk “ The ABC’s and 123’s of Process Improvement.” The organizers hated the title (not academic enough) but a room full of executives had lots of questions and follow-up requests.
In my last post I described the ABC’s of process improvement from that talk:
A - When something that could or should be happening isn’t
B - When something that shouldn’t be happening is
C- When you’re not sure what’s happening
B’s and C’s are respectively, about waste and variation, A’s are about efficacy and opportunity. Focusing only on the B’s and C’s, as is common, can result in immediate but diminishing results. Meanwhile, A opportunities present unlimited upside for organizations.
So, let’s move on to the 123’s for each of the ABC’s - these are the three steps I suggest taking once you identify the type of process challenge you’re facing. While each improvement methodology may have a rich body of knowledge and broad set of tools, these three steps can deliver immediate value and eliminate problems you thought to be baked into your business.
A - When something that could or should be happening isn’t
1. Identify and Exploit the constraint
2. Subordinate other resources to the constraint
3. Evaluate and either Elevate or find the next constraint
Step 1, identifying the constraint begins with a question: “What is most keeping us from achieving more of our goal at this time?” This question has two subtexts - it reminds us to work “goal back” and to focus on one problem at a time. Not every problem creates the same impact - why not line them up?
Don’t be afraid to look both inward and outward for the constraint. Can you deliver enough goods and services to meet demand? Do you need more demand?
Put another way, sometimes the constraint is in operations, sometimes it’s in sales and marketing, product development, or in the supply chain.
The term “Exploit” means to improve the efficiency of a resource or process. This may mean something as simple as using a new tool or as complex as digging into a subprocesses. Don’t be afraid to poke the bear - old policies or performance metrics, well intended when introduced, often cause today’s constraint.
Step 2’s “Subordinate” is another word that might need explaining. Here, I mean it to be the use of other organizational resources to eliminate the constraint, even if it hurts performance metrics in their normal role.
This is a big one. The “divide and conquer” approach to running a business is prevalent, but that approach leads to the optimization of departments or divisions, and ultimately to silos, and not of the organization’s overall performance. There’s a reason Phillip of Macedon coined the term to describe a strategy to destroy an enemy.
What’s awesome about subordination is that these resources are already there, being paid for, and can usually be rapidly engaged.
The last of these three steps, Evaluate and Elevate, simply mean to check and see if the constraint still exists, and if it does, to see if an investment now makes sense. That investment might take the form of a key hire, better software, a well managed buffer of resources, or additional equipment. If the constraint no longer exist, celebrate for a bit and then find the next constraint and get started again with some exploitation! Rinse and repeat.
B - When something that shouldn’t be happening is
1. Walk the process, take and make some friends, draw a map
2. Identify wastes – time (both of staff and clients), resources, information, or materials
3. Use the “5 Whys” to find root causes
These steps generally come from the Toyota Manufacturing System which was Americanized into Lean. In the process they left out most of the fun stuff Toyota executives get to do every day. So I’m putting the fun back in, I’m just not saying it in Japanese.
Step 1 combines three vital concepts - going and seeing, collaboration, and deeper observation. The fun begins with walking the process, talking with the folks you brought and the folks in the process, and noticing the workflows and/or client movements (the people and materials) and also (very importantly) the flows of information.
Many executives try to figure out what’s going on from a spreadsheet when a good walk and a few post-it notes on a wall would tell them much more. So create a map or diagram of the process - it doesn’t have to be fancy, though a few “swim” lanes showing activities as they occur in different groups often helps. While you’re building the map you’ll probably have a few discussions beginning with “We don’t do that!” and ending with “Well, I guess we do.” Don’t worry, this is normal.
Once you have a decent map, move to Step 2, identifying situations where resources or time are wasted, when information is lost or requires reentry, or when we’re doing something that we don’t get paid to do.
I should point out that not all waste is bad - after all, it’s a little excess capacity that allows us to subordinate resources in case A without shutting down another part of the company. And some folks even plan for a bit of “sprint capacity” to allow for a client or project that wasn’t in the original plan. At the end of the day, lean is one thing, anorexic is another.
It’s easy to get excited when waste is identified, but the real payoff is in figuring out why that waste happened and why it’s allowed to continue to happen - what need is being met? Step Three will get you there.
Before we start, the “5 Whys?” is a tool I’ve seen misused so please insert a warning label here: Asking a series of “Why?” questions to “peel the onion” and understand the root cause of a waste has to be done with respect and safety for all involved. “Why?” questions can bring out the defensiveness in anyone. We’re looking for an opportunity to improve, not to pin down a bad decision or a bad person. If you’re facilitating a “5 Whys?” session, no matter how surprised or disappointed you might be, you have to maintain a grateful and productive attitude for the sharing of insights. You just do.
C- When you’re not sure what’s happening
1. Observe the process, chart it if you must
2. Plan, Do, Study, Act
3. Iterate, but only for impact
And finally we’re at C. This is the domain of variation - when our process is to some degree or another, unpredictable or unstable.
I started my career here, doing Statistical Process Control for a silicon wafer manufacturer and I measured and charted the heck out of every possible aspect of the wafer polishing area!
Turns out we’re doing the same thing today in many of Six Sigma projects, Six Sigma being the current incarnation of SPC. I saw one effort in a hospital where over thirty metrics were being tracked, with full statistical analysis, just around the patient’s bed.
One of my first learnings was that I could save a lot of energy and time if I took some time to observe the process, eliminating situations where there wasn’t even correlation much less causation, before designing my data collection effort.
Yes, sometimes the full weight of Six Sigma should be thrown at a problem, but many times a lighter touch is all that is needed. Going all in on statistical methods can be vital, depending upon the industry, but usually not at the first passes in process improvement.
I suggest starting Step 1 with taking time to observe the process and to track inputs to outputs with a few notes. Sometimes collecting data is vital and a graph will come in handy. The reason I say “chart if you must” is that I often see improvement actions delayed in order to collect more, and then more, and then more data. If you’re not impacting safety, move on just as soon as you understand the trends in the data.
Step 2 reflects W. Edwards Deming’s suggestion of changing his famous “Plan, Do, Check, Act” to “Plan, Do, Study, Act.” This was done to emphasize the need to understand results instead of simply validating expectations, and in particular to look for possible unintended effects resulting from any changes made.
While most of it is self explanatory, the most left out part of “Plan, Do, Study, Act” is “Act” - many folks seem unclear on what to do at this point. “Act” means to make positive changes permanent (for future ACTions) through training and workflows. It’s common to see an improvement that doesn’t stick: forgetting “Act” is usually why.
That brings us to Step 3, “Iterate, but only for Impact.” Keep going as long as the process seems to have excessive variance, but consider the cost of each cycle against the realistic value of the results that might be achieved. This honors Deming’s warning that trying to optimize within the natural noise (common variation) of a process can easily result in more noise and worse results than you started with. Or, to quote Bert Lance, “If it ain’t broke, don’t fix it.”
To wrap this up, anytime you get to a Step 3 within an A, B, or C, take a fresh look to make sure you’re still working on the right kind of opportunity. In fact, patterns like A, A, C, B, A are common as you work through one opportunity to the next. Just remember that as you succeed, and the opportunity moves, move with it.
It’s as easy as ABC and 123.