The (hidden) cost of specialization
Published on 4/7/2019

The (hidden) cost of specialization

Why projects usually take three times as long and cost three times as much as planned or are avoided altogether

While the intention of setting up specialized departments is usually an increase in efficiency, managers tend to forget the costs that naturally come with it.

Communication costs

Getting familiar with a business case and its peculiarities takes time. While proper introductions and memos help a great deal in keeping costs low, it still requires substantial communication efforts and continuous learning to stay up-to-date. This is especially true the more complex the project or business is.

Waste as a result of imperfect communication

While so far we assumed communication goes well, normality is the opposite. Misunderstandings are the norm rather than the exception. Until something is successfully implemented it usually takes several iterations and lots of communication. What is saved in communication, usually comes back in doing things more than once.

Coordination costs

While communication overhead is substantial, it gets worse with coordination. It takes a significant amount of time to plan and coordinate — especially with people with a tight schedule. This leads us to the next point.


Things go wrong. Dependencies can never entirely be avoided. Delays in one department affect the other one which leads to further delays. Even though built-in buffers somewhat address this issue, it still leads to unnecessary delays.

Silo thinking and the death of innovation

Since (because of these frictions) cross-department projects can be frustrating, this often leads to one of two outcomes: People either try to do everything themselves or avoid cross-functional initiatives altogether.


Managers have to be very careful when they look for efficiencies through specialized departments. The related costs can quickly be forgotten. In many cases, small autonomous teams with generalists are both quicker and more innovative. And — somewhat counter-intuitively — in the end even more efficient.