Skip to content

Span of control: What is the optimum and what are the drivers?

Published by Rupert Morrison

Span of Control Graph - Icicle Layout - OrgVue

Almost no question is asked more than what the ideal Span of Control (SoC) should be. There are those that believe the answer is pretty simple, 8. You should have no more than 8 layers and no fewer than 8 direct reports. This is easy to understand and I’m sure there are situations where 8 is the ideal.

Our view is it all depends. It depends on a large number of factors and dogma is incredibly dangerous. Believing there is a simple answer is a classic example of a rightsizing magic number.

So, if it all depends, are we just stuck? Or is there a minimum and maximum? The answer is no, no and probably yes (but we can’t say with certainty what that exact maximum is. By the way, the minimum is a bit of a cop out, it is “1”).

Delayering

By increasing the span of control, you reduce the number of managers and layers in the organization. This gives, all things remaining equal, significant cost savings and improves communication from the top.

In our standard OrgVue demo data set, we have an organization with 1,505 people and 229 managers, or an average SoC of 6.6. The person at the top of the organization (depth 1/the CEO) has 6 direct reports. The CEOs reports (depth 2) have an average span of control of 4 while their reports 5.76 and so on, as depicted in the below graph.

Span of Control Graph - CEO Report - OrgVue

The actual layout can be seen in more detail in the “Icicle Layout”. Each direct report is pictured below the boss. (For fun, I have coloured this by age. From Green to Grey).Span of Control Graph - Icicle Layout - OrgVue

Each of the 7 layers can clearly been seen and it is immediately obvious that being at level 7 is quite an outlier and some functions only go to depth 3.

The below graph shows this same data, but a histogram of the average depth by department (with the distribution of the age in colour still there).

Span of Control Graph - Histogram - OrgVue

Understanding this pyramid is useful when thinking through possible delayering savings. It identifies gaps and possible inefficiencies. The simplistic logic of having an average SoC of 8 would decrease the number of managers by 41 to 188. This would mean you only need 5 layers (max depth of 5). This theoretically perfect pyramid would look like the below Excel table.

Span of Control - OrgVue

So, in theory, if you would give us a file with Employee ID, Manager ID (& Employee cost) we would, within seconds, be able to tell you the saving in terms of managers (& likely cost saving/savings range). But reality doesn’t fit such a simple theoretical model. This doesn’t mean such a simple exercise is a waste of time, more that it is useful for highlighting areas to focus in on and a way to raises questions, not answer all those questions.

A framework for thinking through an ideal Span of Control (SoC)

Like most problems, it is useful to break it down into various parts. There is the nature of the subordinates, the boss and the overall context of the organization. Each of these have various drivers. Understanding these drivers should help to answer what the SoC ought to be.

Span of Control - Drivers

Nature of the subordinates:

Nature of the boss:

 Context of the Organization:

It would clearly be possible to define each group of subordinates and bosses into dimensions based on these drivers. From this it is a relatively easy step to deduct which areas should have a higher or lower SoC. High could mean 12 or even 15. Low could mean 5 or even 3 -> with how extreme being determined in part by the organizational context. From this, a gap analysis can be performed and deeper/more meaningful questions asked. Equally, if it is a priority to reduce cost, delayer and increase the average span of control, then why not focus more on the underlying drivers. It is “leveraging the drivers” that will enable an improvement in the SoC ratio. By doing something to improve the drivers (e.g. reduce the amount of admin takes; improve trust or knowledge of managers…) then any possible theoretical saving might just become a reality.