CEO as Success Model Designer

I have just had an interesting conversation with a friend about the state of how things are done in companies, particularly how management models and tools are being used nowadays. I have written and spoken a lot about that in the past, especially their implications in terms of unintended consequences.

Models and tools have a tendency to reduce complex phenomena into a set of things that can be easily grasped by the human mind. While this is exactly how humans perceive the universe, institutionalizing these in a company setting has many downsides. The two obvious ones are: first, the model substitutes good thinking about problems. When you have a two-by-two matrix to be populated, you imply that the two factors are the most important ones. So, even if you have another factor that you think is important for a specific analysis, the model wouldn’t capture it as it is inherently reductionist. Therefore, you may leave out important things from your analysis and end up applying the procrustean bed to the way you analyze things.

Second, models and tools, especially those that are supposed to help managers make better decisions, can actually influence the very same decision, an issue well known in social science called reflexivity. For example, a budget, which is supposed to be used to track results or benchmark against a set of assumptions, tends to influence how people make business decisions. The budget becomes the yardstick of good management and ends up being completely manipulated. Cost centers would want to use all their budgets so that they can get a similar budget and prove along the way that the resources are well spent. P&L managers, despite good prospects, would want to overshoot their top line target and only meet the budget, leaving the rest for the following year.

The other observation I made to my friend is that regardless of what management orthodoxy says, you find real-world examples of companies displaying characteristics opposite to what business success literature would describe as successful, and yet these companies function pretty well. I have even witnessed companies with toxic cultures and unhappy customers, and they are leaders in their industries.

I do acknowledge the fact that there are some general principles for proper management of a company, and most of them are common sense. However, in the absence of a Grand theory of a firm, each company is unique. It has a unique set of people, a unique culture, a unique way of communicating, and so on. The tendency since the industrial revolution is to standardize the way of doing things to capture efficiencies through specialization. And this has worked very well… well, to a point. Today’s increasing complexity, I am not sure this is the way to go.

This uniqueness also applies to the set of factors that underpin a company’s success. In other words, the firm’s success formula is fundamentally unique. This success recipe would be a combination of important factors that differ from one company to another and even change over time. So, the role of the CEO would be to find out what recipe to use or at least get closer to it. I see the role of the CEO as a model designer of his/her company’s success. Still, this model is a reductionist view of the firm, but it’s THE view that matters. As with statistical mechanics, whereby we don’t care about the behavior of each gas molecule but just the overall temperature and pressure, this coarse-grained view of a particular firm is of paramount importance as it helps focus on the few factors that are important and through which success is possible. For company A, it may be a specific set of customers; for company B, it may be culture, despite sometimes being part of the same industry. The model itself is very dynamic as success factors change over time.

The million Euro question is how a CEO can identify what matters. I think this is an attribute that one can cultivate, and I hope that computational technologies can help. Here, I am thinking about an application of Stephen Wolfram’s ruliad, which is the entangled limit of all possible computation where all possible computations are deployed and for all initial states. As computationally bounded observers, we get to parse a very tiny part of it through our sensing apparatus and give us the laws of physics that we know, along with pockets of reducibility in the sea of computational irreducibility that the ruliad represents.

In this computational framework, a CEO embedded in the ruliad can parse his universe (company) and find his own “laws of physics” (success model) and exploit the underlaying pockets of computational reducibility to the company’s benefit.