Some Just “Don’t Get Continuous Improvement (CI)”

I was always dumfounded. I used to say “Continuous Improvement (CI) isn’t rocket science; it just takes a little discipline.” Then a solid fuels scientist told me how “hard” CI was for him. I began to realize that comparing CI to anything rarely worked. The process is deceptively simple.

Measure-Improve- Measure gain, (repeat).

It helps to think in process terms to see tasks in terms:

Inputs — Activities —- Outputs

Still some people didn’t get it. “We do this all the time>” “Ah yes. . . but are you improving?” “Er.”

Then I realized people might think about how they do their work when they first start. They might even track how well they are doing each task then, but soon work is a collection of habits. You “feel” whether it is going well or poorly and adjust.

At the front line work documentation was seen as “busywork.” A point brought home with my favorite line of all time given by a maintenance tech at a chemical plant:

“Paperwork never turned no bolt!”:

At the top level, executives often didn’t view their work (management) as a process and as a result saw the absence a series of tasks -some that workers did and some that they failed to do. When the plant manager heard the maintenanc3e tech above he completely missed the opportunity to show how documentation of maintenance, “paperwork,” allowed him to improve mean time between equipment failure. Instead he sputtered “It’s part of the job!”

I needed a way for people to “feel” improvement, to feel the excitement of practice focused on improvement in a team setting.

The Tennis Circuit Exercise

I hit on this simple but powerful exercise to introduce the concepts of Continuous Improvement (CI) to a team or in an introductory CI workshop. I even used this exercise in an executive overview after which more than 50 percent of the executives signed up for additional training.

In this exercise, participants experience continuous improvement success and failure. As a result, teams understand how process focus and teamwork leads to improvement.

The exercise starts by sharing this objective:

Objective: To create a process and improve it in a sustainable way

Timing:

35-45 minutes. Some groups have such a good time that they will want to keep “playing,” but if the exercise extends to an hour some may lose interest and learning declines.

Required Materials:

One set of three tennis balls for each team (teams of six to nine each work best, but try not to have less than five members per team) plus one stopwatch and flipchart  or whiteboard large enough to make the recorded goal and results visible for each team.

Set up:

Introduce the exercise by saying, “I’m going to give you a chance to experience Continuous Improvement. Please form ___ teams of ___ people each.”

Then show the first instructions:

Create your current-state process – a pattern of how three tennis balls will go through a circuit of each team member, returning to the first person.

You must practice the current-state process at least three times, recording cycle times, before making any process improvements.

Practice each subsequent process improvement at least three times, documenting improvements and times.

Then say,

“Let me ask team 1 to demonstrate.  Form a circle with each person three feet apart.”

Throw the three balls to one person, saying: “Move the balls one at a time from person to person. Remember who you pass to and who you receive from. Okay, have you got that?  Now that is your ‘current-state process.’ Let’s time it.”

Record the time on the flipchart for all to see.

(Note: Throw the balls to the first person. This often sets a pattern of the group throwing the balls between people, which they soon discover isn’t required and limits improvement.)

Continue by saying:  “Okay, now I want each team to practice your current-state process three times, timing it each time before you make any changes.”

Then share the rest of the rules:

• Balls must move independently, i.e., one ball at a time.
• Transfer must take place (i.e., one person should touch each ball one at a time).
• The metric is total cycle time. Cycle time starts when the first ball leaves the first person’s hand and ends when all three balls are back with the original person.
• Everyone must touch the balls in the same sequence in the original pattern of the
  current-state process.
• If a ball touches the floor, or skips a team member in sequence, the circuit must start over with the first ball. The clock continues to run.
• Record current-state cycle time and improved state(s) cycle times on a flip chart.
•  When a change is made, write it down next to your times on the flip chart.

Let the groups compete: 15 minutes

After one or two rounds of changes tell them that the record for this activity with groups of this size is_____ (Pick a number about 50 percent of the average group’s performance to spur the group on. Keep track of records as you conduct this.)

Then after another round read out the times and say, “The fastest I have ever seen is ____”.

Again, pick a number that is 50 percent of average group time to spur the group on.

The actual record I observed was .78 seconds for a seven-person group (a controlled drop) with each ball touching everyone’s hands. You can choose to give this record at the end or just use your own observed records, which is better.

At the end, hand out debrief sheets and ask the teams to fill them in.

Debrief:  15 minutes

Ask groups:

• Starting time, last time, highest time, lowest time. (If the highest time is not the start time, what happened? If the lowest time isn’t the last time, what happened?)
• What changes did you make?
• What waste was removed?
• What factors had an impact (benchmarks, writing down the times, visible metrics, everyone participated, competition, fun, etc.)

What typically happens?

Times usually start out at around ninety to 120 seconds and drop to under ten seconds.

Groups usually figure out quickly that the current-state process is very inefficient if they throw across the center of the group. They typically rearrange themselves in the order of the process (like an assembly line) and move closer.

If a team loses time by dropping a ball and having to start over they may create some kind of “ready signal.” Comment that this is like the pull system called “Kanban” in Lean.

About midway through, teams realize that the rules don’t say the balls must be thrown, just that each person must touch them one at a time in order. This realization leads to creative ways to accomplish that and times drop dramatically.

The debrief

Like any structured experience, most learning comes in the debrief discussion.

The secret is asking questions to connect the groups, e.g., “Did anyone else experience (or try) that?” Did anyone have a different experience?”

At the end of the debrief discussion ask, “What internal processes does this exercise make you think of?  What could you do about that?”

Variations

Goal setting

One variation is to include a goal-setting step after each round. In the beginning, most teams either grossly underestimate or overestimate improvement capability in the next round. Over several rounds the teams tend to get betting at estimating.

Changes to materials:

Bean bags or kushes don’t roll but are harder to catch; the times are about the same. Groups often spend time perfecting the “catch” and fail to move on to creative ways to “touch” the balls.

Baseballs and softballs often start with too hard a throw. Things get broken and/or people can get hurt. Also, people take the balls home to their kids. Using Superballs is just mean. People spend all their time chasing these mega-bouncies and don’t improve much.

A gutter ball is a 1/2” ball bearing and three pieces of cove molding joined to progress the ball bearing from the beginning of one to the end of the third. Good for novelty with those who’ve seen tennis balls, but not as flexible. And improvements tend not to be as dramatic.

Mechanical assemblies. One of the best variations I ever saw was a pressure valve-like assembly that had to be put together with wrenches, micrometers, and feeler gauges. The assembly was used to train maintenance mechanics in a chemical processing plant. The exercise was run the same way except the metric was cycle time and a quality metric arrived at by an “auditor” spot-checking with a micrometer and feeler gauge. Data was entered into Excel, which then produced a graph that showed optimum cycle time below which quality deteriorated rapidly.

Competition vs. cooperation

Often the teams get quite competitive, checking on each other’s cycle time. It’s good to emphasize that you are competing against yourself and draw out the improvement score of the teams in the debrief discussion.

Try stopping midway and give teams an opportunity to share best practices and track high/low and average group time before and after.

Sometimes after debriefing and a thorough discussion of best practices, it’s good to let the teams have “one more go.”  Use the high/low and average group score again and there is almost always an improvement. This shows that there is no plagiarism in CI. So “steal with pride, but share the credit.”

 A simple but effective tool

This exercise isn’t original, some of the positioning and some of the variations are.  It does work. I’ve used it in manufacturing, oil and gas, chemicals, insurance, back office payments in banking. I’ve used it with front line workers to C-suite executives. (In my experience execs are the most resistant to start, but want to keep the exercise going long after the point is made.)

In the chemical plant described above, the maintenance techs themselves  ended up taking responsibility for improving the mean time between failure metric and maintenance documentation ceased to be a problem.

The plant manager said afterwards, “That by itself was worth the fees you charged.” The Tennis Circuit doesn’t get all that credit- it was a five month project, and the maintenance team were stars, but “feeling” continuous improvement was a good start.