Microsoft has been using search conferences for several years, mainly for product planning purposes. Much is known about Microsoft—about the work ethic, youthful energy of its workforce, its fast pace, and prevailing entrepreneurial spirit. Why would the world’s leading hi-tech software developer go head over heals for the world’s most low-tech participative planning method—the search conference?
The reason is simple. It’s the Microsoft culture. One key aspect of the culture is the value the company places on direct, open discussion across boundaries—up, down, and across the organization. The search conference fits Microsoft culture because it’s a practical way for a group to meet face-to-face and agree on an outcome they desire—creative plans for making creative products.
This is the story of one search conference out of several done by the Microsoft Products Groups. The Products Groups build the software that is the core of Microsoft’s business. This search conference is interesting because it is about the practical application of two concepts—the system principle and rationalization of conflict—and how these concepts played out in real life.
The Search Conference
This particular product group is made up of creative people from different fields—producers, editors, film industry people, designers from art schools around the world, and marketing people.
Before the search conference, product plans were in place, but they were not good enough, and the group was being asked to do too many different things at the same time. Similar products were being done in several places in Microsoft and they each had a different strategy. The group was also dealing with a new set of markets to Microsoft that was dynamic. This combined context of ever-changing segmented markets and lack of planning focus caused this product group to look for a new way to do integrated product planning.
The product group looked at several ways to proceed with strategic planning. The one they usually use they call “internally generated from on high.” Microsoft is a company of leaders. It’s a company of people who are there because they are smart and have good ideas. People are expected to say what should happen in their area of responsibility. The way the company culture works is if you have an idea and you are hard driving enough, you are going to get your idea through. In this product group, product decisions were made based on an idea someone had that they had tried out on their own—“internally generated from on high.” The group was looking for a different approach that involved a large group of people and allowed those people to create the strategy they were going to implement.
When the search conference was presented as an option for product planning, the reaction was something like: “I think I get it. Let’s sketch it out.” Then the light went on. “This is great! This is exactly what we are looking for.” They liked the search conference approach because it offered a way to bring people together in a face-to-face meeting to plan and implement their own future.
The purpose statement of the search conference was this: to bring together the creative people from across the organization to develop and implement a plan for the most desirable future of the group’s products.
They got to the part of the search conference where they started discussing the group’s most desirable future. They were looking into the future and clarifying what the desirable futures could be. As they discussed possible futures for the group, conflict happened. People were disagreeing about everything.
Rationalizing Conflict
When conflict arises in a search conference, we rationalize the conflict. We don’t ignore it, diminish it, or placate people. Instead, we take conflict seriously. People work to understand and clarify their real differences. When people disagree we go through a process of making the conflict understandable—slowly, systematically, with the whole group, we identify the fine line between agreement and disagreement.
Rationalization of conflict means making our differences clear and understandable, and it worked like this in the search conference. We hung small group reports on the most desirable future next to each other on a wall, and reported them aloud. Two sets of questions were raised—first, questions of clarification from the groups, and then a question as to whether anybody cannot live with or is not prepared to work towards any item on any of the reports. The rule is: if we disagree on a desirable future point, it goes on the “Disagreed List.” Once on the list, it ceases to be part of the further work of the group. After all, if the people in the room are critical to developing and carrying out the strategy, and even one of them clearly disagrees with a direction, we’d rather know about it sooner than later.
In this process we are not assuming there will or should be consensus. To do so is unrealistic, particularly on topics where there are legitimate adversarial positions. The aim is to precisely establish common ground and to know exactly where the thin line between agreement and disagreement is located.
By the halfway point of the search conference—well into the desirable future—the wall charts with the "disagree list" was loaded with stuff. But the wall charts with the agreed upon desirable future was tiny. As we looked at this we began to ask ourselves: is this group actually a business unit? Is there a strategy here these people are going to be able to get behind collectively? Are these people actually working for the same company? Is this business a system? We began to ask fundamental questions.
A Dilemma—What To Do?
At dinner, the general manager asked Kevin, chuckling a bit nervously: “Gee Kevin, when does the strategy part happen?” We began to realize that in terms of identifying an agreed upon strategy, we were in trouble. Kevin and the manager brainstormed two approaches. Approach number one was to go back into the room following dinner and say: “OK, it sounds like we are having a hard time building agreement around this strategy, so I am going to decide what to do, because we have to build a business here.”
They discussed this and what might play out if that were to take place. They also discussed the fact that they set out to do a participative process and that this approach would not be participative. So they considered the participative alternative. They were not sure, but maybe the group can figure out what to do. Even though she could not see a way through, the general manager trusted the group to come up with the right way forward.
So, following dinner, to this group of people, with the disagree list mounted on the wall, the general manager said: “OK, I’ll admit it, I’m stumped. I have no idea what kind of strategy we are going to build from this. Can anybody help figure this out?”
Normally, in a search conference, disagreements that make it to the disagreed list cease to be a part of the continuing conversation of the group. The group shifts to discovering the common ground they can agree to and make their plans accordingly. But in this search, the disagreements were too fundamental. So we decided to analyze the disagreements to see what we could learn.
The group of thirty-nine began talking. Over the course of the next forty-five minutes, three themes began to emerge from the disagree list. They thought rather than focus on what they don’t agree on, perhaps there are some things on the disagree list that they do agree on. There were three themes and each theme represented an area of agreement. It started to become clear though the discussion that these three themes could be categories people could pursue.
After more dialogue on the three themes, we decided to test the three-theme theory. Kevin said to the group: “Let’s see where we all fall related to those themes in terms of our excitement and energy to pursue them as strategies.” He asked people to physically get up and move to corners of the room identified by the themes, using the criteria of excitement and energy. Everyone moved to one of the spots in the room. As it turned out, three people ended up in one corner. The rest (eighteen and eighteen) evenly split between the other two themes.
We looked at this and said: “This is interesting. What do you think about this distribution?” They began to talk about it and people stated why they were standing in each corner and why that was an energizing strategy and how we could win in that business.
The group of three talked about how they saw their strategy working. As they continued to talk, they identified that, “You know what, what we are saying really fits well with what one of the larger groups is saying over there. We could align ourselves around the kinds of products they are talking about.” So they moved to that group. The search conference continued with two groups now working towards becoming altogether different businesses.
Two System Principles, Not One
In open systems terms, what they discovered in the product group search conference was that they were two systems, not one. It was the open discussion and rationalization of conflict that proved the point.
Each and every system has its own unique “principle.” We call it the “system principle.” A system principle is a clear statement that communicates the organization’s unique purpose, direction, and relationship with its environment. The desirable future and goals that emerge from the search conference contain the elements of the organization’s system principle. If there isn’t one then an organization is not a system. If there is more than one system principle, two or more separate systems may exist within the one organization.
What was so exciting about our experience was it felt at first that there was no strategy, no agreement, and the top person was going to have to suck it all up to themselves and drive her own strategy forward. And who knows what she would have come up with working from the top? She might not have thought of two groups with two separate products. Instead, the manager trusted the group and let the Search conference work its way. The group used its disagreements constructively. By openly clarifying and understanding their fundamental differences they discovered something new. They learned that they were two systems, each needing its own strategy and purpose.
Summary
The search conference has been a valuable method for doing participative planning at Microsoft. Since 1995, there have been more than twenty search conferences resulting in clearly agreed upon, winning strategies. A quote from a business unit manager best characterizes the success of the search method: “When we finished the search, it didn’t seem like we had done anything radically different. Three months later I noticed that people had a deeper understanding of our business strategy than ever before. There were fewer questions about why we were doing the products we were doing. We were much clearer about what we agreed on.”
Postscript: The Three P’s—Purpose, Participants, and Pre-work
We used the usual search conference process. To us at Microsoft, a successful search conference means Purpose, Pre-work, and Participants. We call them the “Three P’s.”
Purpose. Before going into the search conference we found it was critical to get all the participants together to clarify the purpose of the conference. We would wordsmith, craft, and articulate a purpose statement that all forty people could agree with. It was critical because it got everybody “on-board.” It says this is why we are going off-site for three days. One of the other Microsoft search conferences included a 3 and 1/2 hour meeting to craft a purpose statement (a very long meeting by Microsoft standards) on a Tuesday, and then the following Sunday morning brought twenty-nine software developers together to begin a search. That can only be accomplished by a clear and compelling purpose.
Participants. Getting the right people to the conference was critical. Our criteria were simple and straightforward: people with knowledge of the business who are willing and able to be part of the implementation of the plan. This led to selection of a “deep slice” of people up, down, and across the organization. If someone had critical knowledge of the business, but was not part of implementing the plan, we often interviewed them as part of “Pre-work.”
Pre-work. Pre-work meant having participants go out to important stakeholders -- internal and external to Microsoft—before the search conference. They interviewed stakeholders to get their perspectives, expectations, and information about the market. The information was put on what we call at Microsoft a “public folder.” Public folder is an intranet device that allows everyone to review the interview results before the search conference.
Kevin Purcell is an internal consultant at Microsoft. He led the product group search conference. E-mail: kevinpur@microsoft.com
Robert Rehm provides consulting to organizations and communities in participative planning, design, and learning. He is author of the book People in Charge: Creating Self Managing Workplaces, Hawthorn Press, 1999. You can email him at bob@peopleincharge.org.