Management capacity: number of hierarchical layers

In addition to the parameters of direct reports per manager and the absolute number of managers, a key factor that needs to be considered for a company is the number of layers in the organizational structure. Organizations with too many layers can suffer from communication problems because directives must go through a long “word of mouth”—or keyboard-to-keyboard—process, from the top commanding layer down to the lower ones.

Reducing hierarchical layers improves communication and makes the company more agile. Still, the automatic consequence of doing so without reducing the number of employees is to increase the span of control, creating larger teams, which brings other losses. Determining the number of layers and the average number of employees per manager establishes the company’s management capacity.

It would be great if this were a simple calculation, but it is not. There are variations in the seniority level and the number of hierarchical layers, depending on the company’s activity and the strategic position. This is not something that can be imposed evenly throughout the organization. But it’s good to control these parameters with global metrics and metrics by area, so that measures are taken based on data, and not just subjective perceptions.

For instance, a research department working on medium- and long-term innovations may require very senior professionals. Areas such as customer support and internal systems, despite being important, tend to have a higher proportion of junior employees, larger teams, and more junior leaders.

It’s like building a house

A metaphor for the management focus is the construction of a house. While on the one hand management needs to monitor the laying of the foundations and plumbing very closely, on the other, the tiling does not demand as much attention because, should something go wrong, it would be much easier to replace the tiles than to replace the plumbing or correct a faulty foundation. 

Once you define that the average span of control of a company is 10, and that you don’t want more than five layers of hierarchy, the maximum number of employees will be 11,111—1,111 of whom will be managers. Without proper monitoring, it’s quite possible that at some point in the future the leadership will realize that the company has 20,000 employees and only 500 managers. Or that, on the contrary, the number of managers is inflated by 5,000 for a total of 15,000 employees. Various types of imbalances must be avoided or corrected when detected.

The signs that the organization is out of balance will appear clearly in the metrics derived from HR systems, but they can also be clear from other signs:

• Employees go into burnout
• Turnover increases
• Code production declines
• Incidents get more frequent
• It takes longer to mitigate incidents

The DORA (DevOps Research and Assessment) set of metrics can help detect some of these imbalances.

Being data-driven and thinking intentionally about team structure is one of the key features of a technology leader. The organization is always evolving, and the leadership needs to keep a growth mindset, always open to learning new parts of the business and adjusting the team to increase productivity and reduce management overhead. Actively optimizing team structure is part of the platform mindset approach.

Resource allocation

In strategic decisions about organizing a technology area, an increasingly significant variable is the possibility of hiring external services, rather than keeping teams working “in-house.” Outsourcing certain activities is inescapable, especially after the popularization of cloud computing. It is a way of controlling the number of employees, keeping in mind that the company will still need a smaller team to interact with the service provider.

Outsourcing should be considered when deciding a team’s size and structure. It makes sense to reserve resources for developing in-house the systems that comprise the company’s competitive edge and the areas with potential for positive disruption, while hiring services readily available on the market for areas considered as commodities. 

Outsourcing certain activities is inescapable, especially after the popularization of cloud computing.

On the other hand, systems that need to scale could represent a very high resource commitment if contracted out to a third party. It may be more cost-effective to keep their development and operations in-house because this would allow the company to plan these systems for productivity gains and control future costs.