Setting the Limits : Span of Control is NOT About Headcount

What is span of control? You have heard this. Span of control is the number of people you could effectively manage. Essentially it is about how much one person can handle. Because of this reason, there is limited people reporting to a boss. In large organizations, this is one big reason to have 6 to 10 levels from entry level employee to the CEO. There are many levels of senior / top management, essentially doing similar things.

Credit/10pointsmanagement.com

Credit/10pointsmanagement.com

After deep analysis, the HR experts have arrived at number like 10-20 as the right limit for span of control. Perfect. Now the problems should be solved. Or, is it ?

Let us start with two kinds of work  – low complexity and high complexity work -

1. Low Complexity Work:

When span of control theory evolved, appears like there was rather less R&D and more manufacturing. The nature of work had been lot simpler. Once the process and measurements are defined, it is all about monitoring and control. obviously we did not have information technology and right metrics to measure the performance. It involved lot more manual monitoring. Barring the factor of ‘human touch’, one manager could manage many more than 20 people by using tools and techniques available today.

The old school number is no more valid for low complexity work.

2. High Complex Work – R&D, knowledge work:

Here I include all kinds of tech, engineering and other skill based work. You will soon see why the thumb rule of span of control is outdated.

In the context of tech, the technologies in your business, those of suppliers and the customer technologies are changing very quickly. The attrition is increasing across all geographies.  I see that many organizations still go by the thumb rule. To have someone as manager, there are at least 10-15 people.

Okay. For tech we have a manager for 10-15 people. This sounds good. What is the issue?

The real issue is in the perspective. The Span of control is NOT a head count game.

Consider this example..

There is a team of 10 techies. The team is built around the implementation skill –  Python language programming. Python is everywhere. It is used in scripting for animation, complex build automation, test automation in major industries, data analytics and machine learning. Most things related to Google are in python. To use Google apps or any of their cloud, you need Python.

Suppose the team is expected to provide internal services to multiple teams, you will have a manager for 10 members. But these 10 programmers would be working for different internal customers – as diverse as listed above. How do you evaluate the work of each member? How to ensure the knowledge management and proxy planning? These 10 members could be working for 20 different domains/technologies though the language is same.

The key point to note is that the manager needs to understand *some* things about *each of the domains* the team members are working on. That is lot of work for manager to familiarize with. Otherwise, how one would handle the escalations, review the work, provide feedback? Relying on others feedback is not enough.  As I stated ‘Bell curve still exists even after getting rid of it‘. Manager needs to develop the normalized view of all his members.

The above case is a very simple example. Real world scenarios involve complex org structures with miscellaneous topics clubbed under one manager. Many teams/managers do not even have clarity on their internal suppliers and internal customers. I have seen many cases where the customers of the teams are also very diverse. For example – one team may have 1 full product team, mixed with consultancy service in another domain. In this case, there is no good way to define the commonality of the team. For the manager, it is hard to focus on what subject knowledge she needs to develop. If you add the dynamics of organization changes, the team members’ attrition, dynamics of industry and technology, it gets much worse.

As you could see, span of control is about span of topics or span of domains. Lot less about span of head count.

Surprisingly, the organizations give a lot less thought about span of control of topics. It is still seen interms of headcount. No wonder we have project managers who know almost nothing about what team members work on. If businesses think it right, they would spend right budget for senior people to manage even with lot lesser team size. You will have a healthy team with fair work environment.

The managers should manage the work content, not the people. If the work content is managed, everything is taken care.

So, how about creating awareness of span of control of topics? Could we tell the top management and HR that this is no more about headcount?