Sunday, January 06, 2008

Leadership, Diversity and the Goldilocks Zone

By: Gary J. Salton, Ph.D., Chief R&D and CEO
Professional Communications, Inc.

Leaders lead teams. Teams are a tool for getting things done. The leader must decide how to configure a team so that it best meets the needs of the task at hand. One important aspect of design is the size of the team. Another is the particular people chosen as team members. This Research Blog addresses these subjects using hard data rather than opinion and is set in language accessible to anyone.

Actual data was drawn from 3,815 teams. All functions are represented including engineering, IT, finance, medicine and manufacturing among many others. They include profit, non-profit and government organizations. It is a reasonable sample of actual teams operating in the real world. The team size distribution is shown on Chart 1.


Chart 1 shows a “Goldilocks” zone. The area between 4 and 9 people is 25% of the distribution but contains 60% of the teams. This clustering effect suggests that there are forces causing leaders to create teams in the “Goldilocks” range as optimal for their purpose.

The first factor influencing optimal team size is cost. Every added person adds costs. From a leader’s viewpoint, this cost is framed in the difficulty of guiding and coordinating increasing numbers of people. But the cost is not confined to the leader, it is experienced by all concerned. The concept of Transaction Channel is central to its understanding.

A transaction channel is a relation between 2 people in which the actions of one are determined
(at least in part) by the actions of the other. In other words, all parties have to pay attention to what the other is doing, assess it and then decide what to do about it. Communication channels are passive. Transaction channels are active and expensive for all involved.


Transaction channels increase exponentially with team size. Movie 1 shows the process. What starts out as a simple diagram ends up as a bowl of spaghetti. Leaders choosing team sizes that exceed maintenance capacity no longer have a team. They have a group. The advantages of teams will have been lost.


The "right" size of a team depends on the objective being pursued. Leaders set these objectives. They have a choice of pursuing them as a whole or breaking them into pieces. Small teams can pursue narrow objectives. Broad objectives usually demand big teams. If the leader sets too broad an objective excessive cost will have been “built in.” Everybody will pay.

The relation of team size and transaction channels is given by the Combination Formula. The result of applying it to team size is shown in Chart 2. The growth in complexity is geometric.


Firms who sponsor the teams in the database all have a history. It is likely that experience taught them to recognize team sizes that seem to work. The clustering of teams in the “Goldilocks” zone is the expression of this common judgment. At the upper limit it translates to about 40 transaction channels.

There are teams that lie outside the “Goldilocks” zone. But if a leader chooses to exceed the "Goldilocks" limits, there should be a very good reason. The increase in complexity (i.e., cost) is certain. It will be paid by the leader in terms of increased guidance and by the team members in terms of increased complexity. The leader should be confident that the expected value of a larger team size should be enough to offset this certain cost.

Most kinds of diversity are obvious. Age, gender, race and occupational specialization (e.g., engineers, physicians, etc.) are examples. If these are deemed to be important, a diversity mix is easily fashioned. Physical appearance or standard corporate records are usually all that is needed to identify team candidates.

There is one form of diversity that is not obvious—thought diversity. Thought diversity is simply the different ways people can view and address a particular subject. This diversity is embedded in everyone. Engineers can be spontaneous, idea generators or they can be disciplined, methodical juggernauts. Every other category of diversity contains this veiled component. This “secret” form of diversity is what will determine the success of a team.

“I Opt” is a tool that measures the information processing preferences of people. The way we process information determines how we approach issues. For example, if you do not pay attention to detail you will not be precise. It does not matter what you think about it or how hard you try. You have not got the data. On the other hand, focusing on detail limits long-range planning since there are no “facts” about what has not yet occurred. A focus on relationships is a better strategy.

There are more dimensions to thought diversity (see other entries in this blog). However, the above is enough to illustrate the condition. Taking in different kinds of information and processing them in particular ways creates a valid perspective. The mix of these perspectives is what equips (or limits) a team’s ability to get a job done. “I Opt” has defined these thought perspectives as Strategic Styles. They are strategic in that they are a general approach to life's issues. They are styles in that they tend to be applied throughout a person’s life—not just at work but everywhere. They are dependable.

Information processing preferences (strategic styles) can be measured exactly. The maximum diversity any team can have is shown in Chart 3. The area covered can be considered 100%. No matter how many people you add, you cannot exceed these boundaries.


Chart 3 represents a real but theoretical limit. The highest level recorded in the 3,815 teams was 71.2% of that maximum. The reason is simple. If all possible perspectives were brought to bear on a particular issue, nothing would get done. In most real life situations, part of the leader’s job is to set a general direction. This limits the desired thought diversity. For example, if the leader’s goal is replacing a heart valve, there is little value in including people good at expedient action. Including that capacity on the surgical team would be dysfunctional. Diversity is not a universal "good."

Thought diversity is a characteristic of the individual as well as of a team. We all can get ideas, plan, react spontaneously or execute with precision. We tend to favor one or another style, but the capacity for doing everything is within us all. The “I Opt” profile as shown in Chart 4 shows a typical “I Opt” profile. You can note that the profile touches each of the four strategic styles.


The lower theoretical level of individual diversity is zero. However, the only place you would find such people is in mental institutions. In the practice of OD the most typical level will be 19.4% of the theoretical maximum (Chart 3). In other words, a person approaching an issue alone will only cover about 20% of options. This is one of the big reasons we work in teams. Working alone we will probably not see all of the relevant dimensions of an issue.

Adding one person does not double the diversity of a single person. The reason is that profiles overlap. Some portion of the thought diversity the added person brings will already have been covered by the first. This is not bad. This commonality is the basis of mutual understanding. It allows us to work together. This condition is shown on Chart 5.


An example of an actual five-person team used in this study is shown in Chart 6. The outer boundary of all of the overlapping profiles defines the total diversity of the team as a whole. The gray area defines the majority, it is the area where common positions are most easily reached. Both pieces are needed for a team to operate effectively.


The interest of this Blog centers on thought diversity rather than commonality. The outer boundary of the overlaid profiles sets diversity limits. It determines the range of options. For example, if there is no one with an appreciable level of Reactive Stimulator (RS) in the group, the possibility of an expedient solution will probably never arise. If this were relevant (it was not in the surgical example), the team would have been denied access to a valid alternative. Cost, effectiveness or both can be negatively impacted.

The availability of a right level of thought diversity is one of the major benefits of teams versus division of labor work groups. Part of a leader’s responsibility is to create the right level of diversity. Too little and viable options disappear. Too much diversity and the team can be paralyzed. The delays inherent in divergent perspectives can be just as costly as too few options.

The initial 3,815-team dataset are real teams. However, they are a bit of a jumble. They can include transitory teams, “teams” of loosely associated people or teams with a less than well-defined purpose. It was decided use subset of this larger group. The subset includes 1,196 teams where the leader of the group was known. In many cases, the leader’s title was also available. The dataset is still large and does provide increased assurance that the teams were focused on serious purpose.

The diversity index of each team was calculated and set as an increase in diversity over a single person. Thus, 150% indicates that the team increases diversity by about one and a half times. Said another way, the team has the equivalent of 250% of individual diversity.

Movie 2 shows the distribution of diversity indexes by team size. The movie shows both the mid-point (i.e., median) and the distribution (i.e., range) of each size of group considered in this study. It also explains in more detail the forces that affect diversity as it is applied in a team context.


A more abstract but inclusive way of looking at team diversity in relation to size shown in Chart 6. The "Goldilocks" zone generated by team size is superimposed on these diversity levels.


Chart 7 shows that within the "Goldilocks" zone the increase in thought diversity rises rapidly. In other words, every increase in the number of people on the team yields a large increase in the diversity of perspectives. Beyond the "Goldilocks" zone, the diversity value of each additional member decreases. But the cost grows exponentially (see Chart2). The interaction of increasing cost and diminishing diversity value is what causes the "Goldilocks" zone to be created.

Does thought diversity vary by organizational level? Table 1 shows that a common process appears to drive all teams, from President to Supervisor. A diversity index of about 150% over a single person (i.e., 250% total) and a team size of about 9 people appear to be a common standard.


The fact that transaction channels impose costs is beyond question. However, we cannot offer a causal (what causes what and why) formula for why about 40 transaction channels appear to begin to tax the upper limit. Similarly, it is beyond question that different perspectives can contribute to team success. However, why a diversity index of about 150% over an individual seems to signal a level of diminishing value is unknown.

While we lack exact formulas, the findings in this blog can be useful to actual and prospective leaders. Just becoming aware of the nature of the costs and benefits of team design is enough to prevent most catastrophic errors. Being able to attach a measure to both costs and benefits further enhances the probability of success. While not yet complete, this appears to be a useful step forward.