Capelle on Organization Design

A seasoned HR professional has expertise in many areas. However ask them to give a presentation on organisational design and most won’t have much to say.

To address that gap long-time organization design specialist Ron Capelle took the time to put his research, practice and insights into a book.

To cut to the chase, Capelle says that, while there are many important aspects of organization design, the fundamental starting point is position alignment. This includes both horizontal alignment and vertical alignment. Horizontal alignment is the familiar issue of whether we organize by regions, functions, product lines or attempt some kind of matrix organization. Vertical alignment is a concept that is rarely articulated; it is the distance in job size between a manager and his or her direct reports. When I asked Capelle what surprised him in his research he pointed to vertical alignment.

Capelle: The most surprising piece for us was the power of Manager-Direct Report Alignment. It is just one factor in organization design but it had a big impact on outcomes such as financial performance, customer satisfaction and employee engagement. We were surprised at the robustness of that single sub-factor.

Here is a micro-explanation of this alignment. Every employee should have a manager exactly one level above, both in terms of the complexity of the work done and the capability to work at that level. If the manager is in a much bigger role than the direct report, the manager may not be as interested in that person’s work. It can make for a bad relationship; the managers may feel the direct reports are “pulling them down into the weeds”. If the manager is in a role only slightly bigger than the direct report, then they will not have much value to add; the direct report will likely feel micromanaged. The difference in job size has to be ‘just right’.

Requisite organization theory has a method for measuring job size (see time-span) so one can be precise is getting the distance right. Furthermore, job size is not just a continuum of gradually bigger jobs; there are distinct breaks between different “strata” of job size, and the simple rule is that a boss should be exactly one stratum above their direct report.

The easy to understand but overlooked factor of vertical alignment has a major impact on organisational performance.

If you have read the work of Elliott Jaques the concept of vertical alignment will be familiar. Capelle knew Jaques well and has worked to advance the field of requisite organization theory. I asked him whether he had diverged from Jaques’ views.

Capelle: Two areas where we have been consistent with his work are the method of “time span analysis” to measure the complexity of work and what we now call “information processing capability” to measure individual capability. My view is that both of those are scientifically based concepts and we’ve applied them. Our work has added evidence of the validity of those methods; we’ve got 24 original pieces of research and 23 of them are in the book.

Our approach puts a stronger emphasis on systems theory. Elliott’s approach tended to look more inside the organization rather than looking at connections outside. We also put a lot of effort into developing approaches to implement the improved organization design, in essence change management.

Another new element we developed is a method for looking at the complexity of tasks, and we found that professionals spend about half of their time doing lower level tasks that someone being paid less money could do just as well.

The “information processing capability” Capelle refers to is an essential element of requisite organization theory. It mirrors job size in that it does not simply increase in a smooth continuum. There are distinct jumps; and a key to organization performance is to have someone in a role where the size of the job matches their information processing capability (for example as stratum 2 capable person in a stratum 2 position).

Talking about any aspect of individual capability can be tricky. In the West we have an aversion to hierarchy, particularly if we fear it somehow implies one person is better than another. I asked Capelle how he handles this tension.

Capelle: Every organization has a hierarchy. There are different titles, different levels of compensation and different reporting relationships. We don’t create this, we simply recognize that it is there. There is a tendency to pretend organizations are not hierarchical rather than saying we do have hierarchies and so how can we make them as effective as possible.

Similarly, organizations already make judgments about individual capability. It is built into recruiting. It is built into matching people to positions. It is built into promotions. Our use of information processing capability does not increase the level of differentiation. However, it does provide a method to help make better decisions.

We need to be careful because information processing capability is just one of three things that are important in matching someone to a job—the other two being “skilled knowledge” and “application” (i.e. valuing the work and fully applying oneself to it). Measuring information processing capability makes assessment more precise, it allows managers to make better judgments then they would otherwise. It is not a substitute for managerial judgments, but it is a method to help people make better judgments. An objective is to help people use their full capabilities and be successful… neither over-promoting to the point of failure (Peter principle) or holding people back from using full capability.  

This is not about some people being better than other people. This is about the levels of work that people are best suited for. Organizations need to be careful anyhow about not implying that because you are higher up in the organization somehow you are better. The reality is that all people and levels of work in an organization are critical, and for an organization to be effective it’s necessary to pull everything together.

The other thing that is important is that information processing capability is something that can grow over a period of time, so if someone is at a certain level it doesn’t mean that they are necessarily going to stay there all of their lives.

Capelle also buries the unhelpful myth about span of control.

Capelle: Over the years we have developed a database of over 59,000 manager-direct report relationships from over 76 organizations. One of the things you sometimes hear is there is a single ideal span of control. This goes back in the management literature a long ways. A number of consulting firms build this into their practice.

We’ve never thought this was correct. We believed that the correct span of control was related to a number of different variables at a given point in time, and could therefore be quite different in different situations. Now, if there were a single ideal span of control then one would expect it would be related to better employee satisfaction. Our research showed no relationship between a single span of control and employee satisfaction. This raises serious questions about their being a single ideal span of control.

An important element of organization design that is often overlooked is the notion of a cross-over point manager: the point of accountability for coordinating silos. I asked Capelle to elaborate.

Capelle: The cross-over point manager is the one who has most or all of the resources necessary to deal with an opportunity or issue. Identifying the cross-over person, and making that accountability explicit, doesn’t mean that the manager has to do more work. The cross-over point manager on an issue sets context and prescribe limits for what should happen; that doesn’t have to take much time.

Being clear about the cross-over point is particularly helpful in IT projects. Too often, the CEO does not take the time to talk about what the different accountabilities and authorities are with respect to other affected departments. When the IT system doesn’t work the way it is supposed to you get finger pointing: the business unit people say that IT didn’t do a good job and IT says it didn’t get the support it needed from the business.

If the CEO, being the cross-over manager in this case, does relatively minimal work in clarifying accountabilities and authorities then they may not have to get involved at a later point to clear up a mess. I’ve seen a SAP project that was floundering and we were able to identify what needed to happen.  The CEO got the people together, set the context and prescribed limits, and talked about the broad accountabilities and authorities. Beyond that the CEO had very little direct involvement because things were clear in terms of who needed to do what.


Ron Capelle’s book is called Optimizing Organization Design: A Proven Approach to Enhance Financial Performance, Customer Satisfaction and Employee Engagement. It is available at booksellers.