Learning Histories: A New Tool For Turning Organizational Experience Into Action

by Art Kleiner and George Roth

Editor: Suzy Wetlaufer

Date: June 3, 1997

In our personal lives, "experience is often the best teacher" or so the old saying goes. Why not, then, in corporate life? After a major event , a new product failure, a wild business breakthrough, a downsizing crisis, or a merger , many companies seem to stumble on in ways that miss the lessons of the past. Mistakes get repeated, while smart decisions do not. Most importantly, the old ways of thinking which led to past mistakes are never talked about; which often means that they are left in place to spawn new mishaps ad infinitum.

Ask individuals within a company about such major events, however, and they will often tell you they understand exactly what went wrong (or right). You might hear that the new product fizzled because no one in marketing listened to the people in manufacturing , or vice versa. Or the new product soared largely because R&D, or distribution, "finally got their act together." Each of these points of view represents a valid, but limited, piece of the answer to the puzzle of what happened and why. If all these perspectives could somehow be melded together coherently, the organization as a whole might learn what happened, why it happened, and what to do next.

Yet these kinds of insights about organizational experience are rarely shared openly, and they are analyzed, debated, and ultimately internalized by the whole organization even less frequently. In other words, in corporate life, even when experience is a good teacher, it's only a private tutor! People in organizations may act collectively, but they only learn individually. This is the central tenet , and frustration , of organizational learning today.

The frustration exists because managers have so few tools , if any , with which to capture and disseminate their institutional experience. Employee surveys are often used to gather information and opinions about major events that have shaken up a business, but the assembled data rarely makes it back to the people of the organization in a form they can use meaningfully. "Best-practice" write-ups leave out the mistakes and mishaps which people might learn from, as well as the hidden reasoning and struggling which made visible breakthroughs possible. Sometimes consultants are called in to make sense of the "big-something" that has happened. But their reports are rarely embraced by those who lived through the experience. The reports, after all, are aimed at the senior managers who hired the consultants. Once the consultants leave, the lessons of the past slip away with them, often to be sold, to other companies.

A Different Approach to Institutional Learning

How, then, can organizations reflect collectively on past experience , and do so in such a way that people's thinking and acting are more focused and energized in the future? In other words, how can the lessons of the past be "processed" by an organization so that they translate into more effective action?

Galvanized by these questions, a group of social scientists, business managers and journalists at MIT's Center for Organizational Learning have spent the past four years developing and testing a tool to solve the conundrum of collective learning. We call our solution a "Learning History." (For an annotated illustration of a Learning History, see the pull-out accompanying this article.) In the most basic terms, a Learning History is a narrative of a company's recent set of "critical episodes" , a corporate change event, a new initiative, a widespread innovation, a successful product launch, or even a traumatic event like a downsizing. The document ranges in length from 25 to 100 pages, nearly all of it presented in two columns. In the right hand column, relevant events are described by the people who took part in them, were affected by them, or observed them close-up. Managers, factory line workers, secretaries, and outsiders such as customers, advertising copy writers, or suppliers, tell their part of the tale. Each person is quoted directly, and identified only by title. The words are woven into an emotionally rich, coherent story that reminds some people of Studs Terkel's well-known, unvarnished first-person accounts of American life and society.

The left hand column is a different matter. It contains analysis and commentary by the "learning historians" , a small team comprised of trained outsiders, usually consultants and academics who specialize in organizational learning, along with concerned and knowledgeable insiders, usually drawn from the company's human resources or organization effectiveness departments. This team has sorted through hundreds of hours of interviews to "distill" the story in the right-hand column. And in the process, they have come up with the text for the left-hand column, which identifies recurrent themes in the narrative, poses questions about its assumptions and implications, and raises "undiscussable" issues that hover just below the surface of the quotations along-side.

A learning history is not completed just to be distributed wholesale through the organization and read , or, more often, shelved , by individuals. Instead, it is used as the basis for group discussions for those involved in the event, and those who might learn from it. (This audience might well include every manager and employee in the organization). For instance, a learning history about one division's successful product roll-out may be used to spark conversations in another division about to launch its own new product. The members of the latter division are asked to read the learning history, marking portions of the text that upset, excite, or otherwise engage them. The learning historians then meet with these people in small groups, facilitating an open dialogue about the ways of thinking that led to the first group's success. The goal of these meetings is to get a better understanding of the critical choices faced in planning new actions. In this way, a learning history is as much a process as it is a product.

As for the product, its "jointly-told tale" format may seem unique, but it is in fact based on an ancient practice: community story-telling. Since the beginning of civilization, tribal peoples have gathered together, perhaps around a fire, to retell stories of important events , wars, changes in leadership, or natural disasters. Many individuals offer their remembrances (or what we would call "perspectives") during these gatherings, and throughout, a shaman , the learning historian , comments on the narrative, guiding a community discovery process that brings the significance of the story into the light. The group hears a multi-faceted tale, but with one directed purpose. They have re-experienced an event together, and learned collectively of its meaning. Indeed, the group has created this meaning together.

Are we suggesting that this timeless story-telling form will prove powerful in a corporate setting? Preliminary results indicate that the answer is yes. To date, more than 15 learning history projects have been conducted, mainly at large American companies trying to make sense of major, controversial incidents in their recent history. In one case, a car company new-product launch team broke internal records for quality and speed to market; the learning history illuminated the new kinds of cross-functional interrelationships which had led to those results. In another, the learning history examined the transformation effort at a Fortune 50 company, in which many entrenched business units were eliminated, others combined, and several new ones introduced. Fall-out from the transformation left thousands of employees struggling with questions about the culture of the new organization, and the role of managers within it. The learning history helped many people move forward by showing the common, unspoken dilemmas that the entire organization was wrestling with: How to act entrepreneurially, for instance, amidst a bureaucratic legacy.

Why Learning Histories Work

In general, we have observed that learning histories have several effects. First, and perhaps most important, they build trust. People who believe their opinions have been ignored in the past come to feel validated by the presence of those opinions in the document (no matter who expressed them). People who have felt isolated come to feel as if they are not alone in their efforts to carve a better future for themselves and the company. Finally, the small group discussions that accompany the learning history provide new opportunities for collective reflection. These help people clear the air about their own concerns, fears, and assumptions, and thus develop a higher level of confidence in each other. As trust grows, it creates an environment more conducive to learning, particularly collective learning , because collective learning depends upon the candid sharing of ideas.

Second, learning histories appear to be especially effective in raising issues that people want to talk about, but have not had the courage to speak out loud. The document, with its anonymous right-hand column commentary from participants and pointed prompting in the left-hand column, does this for them, in a safe environment. Thus, a long-standing rivalry between two plants in a manufacturing company came to surface in one learning history, in a way which showed how both sides had, in effect, colluded to keep that rivalry going , at the expense of the quality of the machines they produced together.

Third, a learning history has proven particularly successful at transferring knowledge from one part of a company to another. Instead of merely copying others' "lessons learned" (which may not fit their situation), readers of the learning histories can read about the reasoning and impulses which had led to those lessons, and apply the insights to their own implementation.

Consider the case of a learning history which took place at an oil refinery in the Midwest U.S. For several years the plant managers and employees had sought better ways to control operating costs. A breakthrough occurred when a new, cross-functional action team decided to "get to the bottom" of a chronic problem with an overheating compressor. In the course of collaboratively discussing, planning and executing a solution, the team developed a new maintenance strategy, that helped solve a wide range of equipment problems - and that ultimately saved the refinery $1.5 million. This single solution wasn't the focus of the learning history. Instead, the 20-page document was distributed to the rest of the plant's 600 employees, and at a number of refineries worldwide, to demonstrate that innovative solutions could be forged internally - and how to start the necessary conversations. "The learning history was extremely important to our proactive manufacturing effort," says one manager at the oil refinery today. "It was a way for everyone - operators and managers alike - to recognize that a more proactive manufacturing approach had a shot, and maybe they should help it. For the rest of 1995 and 1996, we referred back to the story at key moments. And we generated 50 more such innovations at the refinery."

Finally, learning histories also help build a body of generalizable knowledge about management , what works and what doesn't. Learning histories are commissioned to analyze one event, but their lessons often supersede it. For example, one recurring lesson is that "hard" results, such as financial returns or technical objectives, frequently depend on "soft" issues, such as company culture and the level of trust within an organization. Indeed, the learning histories to date have strongly suggested that in reengineering, redesign, or quality efforts, the single most critical factor for success is the quality of human interaction in the organization -- which in turn depends on the humility and openness of the leaders who direct the effort. There are other recurring themes in learning histories, too , so many, in fact, that someday learning histories may be routinely included among the textbooks and treatises in business schools and libraries, to be used as an important source of insight for those engaged in developing the science of management.

The Future of Learning Histories

Without question, the learning history tool is emerging from its experimental stage. We will know more about its effectiveness in several years time, when we can revisit companies that have been through the learning history process. Have the lessons of this newfangled document , rooted in an age-old tradition , continued to have an impact within the company? Only experience will tell. After all, experience can be a great teacher, in both our individual and organizational lives.

The following excerpt is drawn from a 1996 learning history describing the very successful launch of a new car , here with the disguised name of the "Epsilon" , by a large automobile company, which we will call "AutoCo." The authors of this article were originally called in to create a learning history about the Epsilon project in order to help other vehicle-development teams within AutoCo learn from and improve upon Epsilon's process and managerial innovations. For instance, the Epsilon hit the market one week ahead of schedule , virtually unheard of in the car industry , and spent only $15 million of $65 million budgeted for last-minute changes to the car's design. What had gone right, senior management wanted to know, and could the rest of the company experience the process vicariously, as it were?

For this particular document, the learning history team consisted of the authors, plus three members of AutoCo's training and development group. We spent three months interviewing 45 people connected with the Epsilon project from engineers to secretaries, and up the ladder through AutoCo management. We then spent three more months sifting through the thousands of pages of interview transcripts for meaningful, representative quotes, constructing the most relevant narrative story line, distilling central themes to be illuminated in the left-hand column, and basically putting the 89-page "book" together.

The following vignette was included in the learning history because it so perfectly captured a central lesson of the Epsilon project. The new car launch succeeded, in large part, because Epsilon people continuously broke , and then re-invented , AutoCo rules and procedures. But this up-ending of business-as-usual also created conflict and confusion within the organization. Listen as those who lived the experience tell their divergent versions of what happened.

The AutoCo learning history contained significant messages for AutoCo. If breaking AutoCo rules was one of the reasons Epsilon succeeded, for instance, what did that say about the rules? For this reason, the learning history was not disseminated to the organization until senior management had read, discussed, and accepted it. Indeed, a top executive ended up writing the document's introduction, stating that management acknowledged the story's lessons.

In the year that followed, the learning history was discussed by hundreds of AutoCo employees in small groups, facilitated by internal consultants. It continues to be used within the organization today, as new teams embark on the challenging journey of new-product creation.