What are the most important methodological problems that are obstacles to the automation of managerial work?

Posted by

In 2017, on the basis of previous studies of managerial behavior with the help of online managerial tools on the TransistorsHead.com platform, I published an article in which I described the 5 most important such problems hindering the development of managerial automation. I would even say that they are hindering the development of management science and, by extension, the development of business.

I will briefly outline these problems here, and you can find the entire article below.

First, the methodological mess

This problem is that in the management sciences it is not possible to apply the method of incremental knowledge about a certain slice of reality that is typical, for example, in the sciences. Research is carried out on hundreds of topics, within the framework of thousands of concepts and theories (at least that’s what these creations are called, because they are usually not theories). Research results are hardly comparable. In a word – chaos. This chaos has been increasing for more than 100 years.

Second, the representation of the manager’s work

For many years, management science has used three theoretical concepts by which to describe the work of a manager. These are: managerial roles, managerial skills and management styles. I have written about this quite extensively in several posts, for example here:


None of these concepts provide a way to know what a manager actually does.

Third, research methods

In management science, the survey or interview method is usually used. Very rarely the observation method is used, and even more rarely the experimental method. Unfortunately, but the survey method cannot get an answer to the main question in automating the work of a manager: what does a manager really do? And without answering this question, we will never build a robotic manager.

Fourth, measurement tools

The traditional measurement tool in management science is the survey or interview questionnaire. Like the corresponding research methods, it does not provide an opportunity to gather facts about the manager’s work, but only opinions, and to this cited after some time. How human memory works and what managers remember, read here:


Fifth, inference about reality

First, let’s establish the only inference apparatus we have that leads to true conclusions: formal logic. It is very rarely used in management science, and the most common is incomplete induction, which allows us to infer the full population on the basis of n cases. Does it really allow and does it give true conclusions? This question I leave unanswered…

I present the entire article on the problems and proposals to solve them here:

Flak, O., Selected Problems in Team Management Automation