Excerpt from a Learning History:

The Epsilon Project is "Worse than Red!"

As they practiced their new management approach, the Epsilon product launch team came to feel isolated from the rest of the AutoCo culture. They felt that senior managers sometimes applauded them, sometimes supported them, sometimes ignored them, and sometimes invalidated their efforts. Meanwhile, some senior leaders perceived the Epsilon team as withdrawing into its own "true believer" approach as if Epsilon leaders felt that they knew how to achieve results that the rest of the AutoCo organization did not. This problem came to a head around "change requests" (CRs) the engineering reports of issues, problems and impending changes on a part which reached a count of 524, at a time when a more typical car launch would have a high of 200.

Each short story within the learning history begins with a title meant to engage, or maybe even incite, the reader, just like the title of a new novel or movie. The operating assumption is that learning is a pull process: participants need to be drawn into an active intellectual role -- as opposed to a push process, in which participants remain passive, while "lessons" are pressed upon them.

The vignettes themselves begin with a full-column prologue, which sets the stage for the episode that follows. In this case, readers learn that Epsilon and AutoCo perpetually disagreed about change requests (CRs) -- official reports issued by engineers alerting the system that changes in the car's design will be necessary. A learning history's prologue is invariably based on business outcomes or facts that no one would dispute. This signals to readers that the episode is worthy of their attention.


Is it required that other groups share the perception of what a metric means and how it is used? What does this require of groups who seek to innovate processes?

Engineer: In the past, engineers would keep a hidden log of their problems. We would make them public only when we knew the answer. To say we were not rewarded for revealing CRs would be an understatement.

Typically, more than one person would be trying to solve the same problem. We would not know what each other was doing, because there was no common document that tracked the problem. I might be working to solve something involving sheet metal. The sheet metal people wouldn't know because I didn't make it public on a CR out to the world. I might not even have known it affected sheet metal. When I finally wrote up my solution, they might say: "Wait a minute. We can't do this."

With our new process we were encouraged to get CRs out in the open sooner. Now everyone knew each other's problems in time to work together on a solution. This meant that the Program Manager had to empower us to handle our own problems.



In the right-hand column, participants (identified only by title or position) tell their part of the story. We select the quotes because they move the story forward dramatically, but also because they represent important perspectives on the episode. Here an engineer describes his version of the events, while revealing the underlying assumptions and attitudes that made him and other Epsilon engineers challenge the traditional CR protocol.

Here, a manager outside the Epsilon team expresses his dismay over the visible new signs of stress.

What made it difficult for Epsilon managers to explain their strategy and behaviors in promoting early reporting of concerns?

 Senior manager: In manufacturing operations we have a metric that starts with green and goes to yellow and goes to red. The Epsilon program was "purple," I said. That's worse than red! I've had a unique ability to say what's ready or not, and be right 95% of the time. The Epsilon wasn't ready. The patient was terminal. I recommended postponing the launch date.

Epsilon team manager: Initially, I felt really good about our CR count. It was fantastic to find out about all these things and have them worked on. "By encouraging engineers to write concerns," I said to our vice president, "we're actually getting work done earlier and we'll have a better-quality product. This is a change in our system and we want to keep it that way. We want it to be not punitive for an engineer to write a concern early." [The vice president] nodded and listened. But after the meeting, he still said the program was out of control.

The left-hand column provides commentary, insights, questions, reflections, and perspectives brought forward by the learning historians. These comments are designed to provoke deeper, more reflective conversations as readers gather in small groups to talk about the implications of the company's "big event."

To tell a story from multiple perspectives requires bringing in quotes from a variety of people who build and elaborate the story in counterpoint to each other. In this particular learning history, initial interviews were conducted with Epsilon team members. As they talked about the reactions of other AutoCo managers outside the project, it became necessary to interview some of those managers, such as this vice president, to make sure their point of view was included, and treated fairly.

 Vice President's perspective on the same incident is at right. To him, the Epsilon project was not "out of control": it was simply going through the normal expansion and retraction of changes.

Vice President: The ethic that everybody was trying to follow was: There is a right time for change and there is a wrong time. Let us optimize the product and process so we have a quality launch. That's the way it's supposed to work. It never does work that way. The engineers keep changing things, most of the time for good and relevant reasons. Manufacturing drives some changes. And so you have this constant battle of late failures, problems, fits, and finishes. There is huge pressure from Manufacturing to drive the change count down.
This comment draws the reader's attention to the implicit meaning of the Vice President's quote. In effect, he was disagreeing with the Epsilon team's view about CRs, but not saying so directly. These types of unspoken disagreements are typical in any major change event within an organization. A learning history can surface these "silent" conflicts, and make them rich fodder for discussion and debate. And in doing so, the learning history helps members of an organization understand, and often appreciate, why different perspectives exist. The goal is that in the future, people may then be more open to "opposing" ideas.

 One year later, under pressure over the high number of change requests, the Program Manager and the Launch Manager instituted a change in procedures. Engineers were told to stop everything else and resolve changes. During an intensive weekend, the engineers reduced the number of open CRs from 350 to 50. At the time, this enhanced the program's reputation. The Program Manager had demonstrated that he was in firm control.

Ironically, however, the appearance of solving problems early may have contributed to an outbreak of late-breaking problems when changes that had been "pushed underground" resurfaced later in the game.

 A full-column text interlude is used when a transition is needed, like an emcee announcing a shift between stage acts. This text sets the context for later quotes as members of the organization continue to reflect in this case, upon the implications of Epsilon's compliant return to the "old system" for handling CRs.

 What sort of agreement or "buy-in" is appropriate to ask for from other organizations and non-team members ahead of time?

Engineer: But when management takes that approach you drive your engineers underground. Nobody will write a CR that they don't have a solution for if they know that their supervisor has been told to come to them three times a week to ask them about their open CRs. The engineer won't tell me about it. Thus, after we got through the VP build it reverted back to the old "hidden" system.  

 In physics there is a law which specifies that any action is met by an equal and opposite reaction. Does this law from the physical sciences also apply to the behavioral arena of managerial action? What would be your reaction to pressures to reduce CRs?

Epsilon team manager: From then on, we tried to talk informally about our concerns. But that's not what you really want.
You really want a formal system in place that lets everyone know when there is an issue. Once the whole company system knows a concern exists and it's a problem, they can all think about, "Now, how might that affect me?" Everybody can work on it together.
I even went back to [the Vice President] and said, "The magic of this system is we capture everything, I mean everybody knows about it from the day we capture it."

He thought that sounded terrific, but he still didn't like open concerns!

Learning histories can help companies make sense of a particular experience, but they are also designed to bring their readers up to 20,000 feet -- to see that one experience embodies larger, generalizable lessons. Here, a law of physics is evoked to spark the insight that seemingly small changes can jolt the whole system.