Meet Emil and Marc. Emil just signed a contract to work for Marc. This makes Emil an employee and Marc a manager. With his signature, Emil has agreed to follow the orders of Marc. Disobedience is an option, but it comes with the risks of being fired.

Marc the manager points Eric to chop a stack of wood. By doing this Marc is using the most basic form of a management practice, the direct order. Next day, Marc orders Eric to stack the firewood on a need pile in that corner over there. On the third day, Marc is late. Eric sees a stack of wood, and being human and not an automaton, starts to chop it, like on day one. Without knowing, Eric has developed a job description for himself: “My job is to chop wood and staple it”. The job description is another basic form of a management practice. It spares Marc the Manager the time and effort to direct Eric. Unlike a robot Eric the Employee is able to see the work and do it, without being ordered. Marc may continue to supervise Eric, but he might find a better use of his time in carting the firewood to the market and sell it.

One day, after a heavy rainfall, Eric sees that the roof of the shack, where the firewood is stored, needs repairs. Without being ordered, he fixes the roof. What Eric did is to use his judgment of Marc’s interest and decided to act autonomously. Marc has not directed Eric to do that, but Eric has developed a sense of purpose in his work, and chances are that he feels responsible for it. Marcs comes back later in the day and wonders that Eric has not produced his usual stack size of firewood, but he sees that the shack is repaired. Marc may tell off Eric for not making the numbers, but he decides to praise Eric for having taken the initiative and prioritizing repairing the shack over his chopping duties. Thereby Marc has embraced another two basic management practices: Feedback and Delegation. Eric is no longer just following orders but he is empowered to do other things necessary to keep up the production of firewood.

Why has Marc opted to praise Eric and accept his autonomous acting? Marc, hard pressed to make living out of his business, see’s those management practices as being efficient. In his mind, Eric has saved him a lot of trouble, as wet firewood doesn’t sell. Marc may not know it, but he has developed the performance hypothesis in his mind that Job Descriptions, Feedback, and Delegation produce better results, than just ordering Eric the Employee around. Marc the Manager benefits from adopting those Management practices. Eric the employee likes being responsible, too, which is part of why these management practices are working. But even if Marc didn’t give a damn about Eric, he knows he would hurt himself by not employing these practices.

Over time Marc might decide to adopt other management practices, like

  • a regular, weekly meeting to discuss issues
  • providing a budget to Marc that he can spend on axes or saws
  • a bonus scheme based on Erics productivity
  • job sharing, so that Eric is assisting Marc at the market from time to time, in order to get a larger picture of his duties and exposure to customers
  • Annual objective setting and performance review  to clarify high-level targets for Erics work

Marc the manager will introduce and maintain these management practices only if he expects that these contribute to the performance. Margins in the firewood business are so slim these days.

The Case for Constant Experimentation with Management Practices

Shouldn’t any company seek to emulate Marc’s way of working? Things like…

  • Adding new management practices if they work
  • Getting rid of those that don’t seem to work
  • Constantly adapting practices to the need of the business

In a business world that is ever-changing, why do we emphasize so much the need to act like a daring entrepreneur, who finds ever better problem-solution fits, but overwhelmingly fail to engage in experiments with the very ways we are working together? Instead of seeking to constantly improve our way of collaborating with one another, we focus hard on business models, productivity figures, financial performance.

Marc would see that fixation with direct business results as being silly. Results are important, yes, but they can not be enforced directly. Instead, they need to be approached obliquely, by working better together. If we can achieve that, results are not guaranteed, but they will come much more easily.

What is a business if not a sum of decisions taken at all levels of the company? If we can just increase the quality of decisions by some minuscule percentage point, isn’t a companies performance bound to increase? Better management practices result in better decisions result in better performance.

Management practices are like the underlying factors of a companies performance formula.

  • Company Performance = f (Strategy, Execution, Chance)
  • Strategy and Execution = f (Management Practices, Chance)

In other words management practices, the way work in done, influence a companies ability to come up with a good direction (strategy) and competent implementation (execution).

This sounds like a no-brainer. But there are three caveats with this logic:

  1. Managers do not care too much about the performance of management practices
  2. Owners care about performance, but can’t really observe the impact of management practices on performance
  3. The empiric, scientific evidence of the link between management practices and company performance is weak

Manager’s Do Not Care so Much About Performant Management Practices

Marc the manager holds four distinct advantages over most other managers:

  1. Direct Feedback: The impact that the management practices he adopts have on Eric’s performance are very direct
  2. Underlying simplicity: The firewood business is simple. Causes and effects are directly visible
  3. Small numbers: It’s just Eric the employee, not a group of employees or a host of departments to coordinate. This spares Marc the manager from the otherwise inevitable power and social dynamics
  4. No agency problem: Marc is the owner and the manager. He is able to prioritize performance of the business very highly – his performance and the business’s performance are the same. Managers, who are not owners, quickly see their well being and the businesses well being as two separate things
  5. No ingrained, legacy practices: Most managers join companies that have a certain way to do things, a certain management culture. It’s much harder to experiment with management practices if social norms are already firmly entrenched

For a typical modern-day manager, it is not only much harder to see whether his way of managing works better than other ways. On top of that, an employed manager does not even share the same passion for performance than an owner. Risk minimization by not sticking out one’s neck, social conformity and self-optimization might be more important than performance optimization. The fact that the performance of one’s management practices employed can’t be measured easily compounds this agency problem.

The result is that performance becomes a secondary concern while selecting management practices. Control is much more important.

Owners Can’t Really Tell What Management Practices Work

Owners care about performant management practices, don’t they? After all, it is their money that is wasted. But even owners care for performant management practices is limited:

  • Ownership might be diluted. If an ownership share is sufficiently small, influence is very limited.
  • The Agency problem, again: Managers, who are in day to day contact with the business know a lot more about the business they are managing than owners. Owners might employ a few checks on managerial powers here and there, but finally, owners have no option, but to trust.
  • There are other factors easily observable, like those found in the P&L or balance sheet. By their very nature management practices do not lend themselves to be measured in hard numbers. Humankind is excellent at measuring financial systems, but we suck at measuring social systems

The point that I am making is not that no one is not concerned with the performance of organizations. Indeed, there are many people caring about profits and corporate outlooks. The point I am trying to make is:  Few people are making a major effort to influence the performance of an organization by virtue of its management practices.

Science found a bit of evidence, just a bit

Financial performance is a primary concern for any company. But it is usually tackled head-on by looking at market share, product portfolio, customer bases, competition, cost structures, distribution networks, business models etc. Management practices get into view only with hindsight: If a company is successful, it must have great management practices. Phil Rosenzweig, a professor at IMD in Lausanne,  has written a whole book about the ex-post sanctioning of management practices. He named this the “Halo Effect”. Huge business books bestsellers like Jim Collins “Good to Great” or its predecessor “Built to Last” or Robert Watermans “In Search of Excellence” fell for the Halo effect. Great stories, but no scientific value.

But there are a few recent studies that imply a link between good management practices and a companies performance. According to one of those (Bloom et al 2011)  management practices explain about 10% of the success of companies. And according to another study (Bloom, Mahajan, McKenzie 2011) that link is causal, i.e. management practices improved first, company results followed.

That is not overwhelmingly strong evidence. But this is only natural: We just can’t measure social matters with the same exactness as physics. Social systems are highly idiosyncratic things. Take for example the human invention of the stock market. The way prices on the stock market are determined is a result of the human social system, the value humans attach to the stocks listed. Despite hundreds of billions of investment, no one can predict stock values with any certainty. Great efforts are being made in analyzing stocks, but finally, all this effort is undermined because we suck at measuring social systems. It hard to predict human behavior with certainty. Social systems are even more complex than the individual human actor, so science is bound to fail. There are no social physics, no immutable rules. There are things that appear to work for a time, but that is no guarantee that those correlations will hold in the future.

Experimental Management

To sum up my argument:

  1. There is a clear logical link between management practice and a companies performance. The sum of all decisions of all employees should make a great deal of difference to a companies success.
  2. There is academic evidence of this link, but it is weak
  3. Owners and Managers prefer management practices that work over optimal management practices. All sing the hymn of performance, but asymmetrical information, the pure opacity of causes and effects in social systems and individual incentives let them focus on the observable, largely financial facts, instead of the underlying intangible social performance of the organization

My point is:

  • If we can’t say what management practice is really working, why are nearly all companies keeping their management practices static?
  • Do such companies suppose they already found the optimum?
  • Those companies implicitly assume that there is nothing to gain from experimenting with management practices
  • Is it not silly that Lean Start-ups, Entrepreneurial and Agile Movements all have a strong emphasis on experimentation, but experimentation with management practices are of a (at best) secondary concern for most companies trying to become fit for the Digital Age?

Therefore, I suggest Experimental Management. If we don’t know what works best at that time, we need to try things, observe the effects and tune and tune and tune our way of “doing things in a group”, of managing.

We do not need big theories of Leadership and Management for this. We just need to experiment and watch. In other words, managers need to work empirically, not ideologically. Find out what works themselves and not following snake oil selling business book authors, leadership gurus or opinionated non-empirically focused consultants.


Experimental Management is a term that is slightly provocative to our cultural norms. First, we expect competent management that knows which practice works. Dabbling in management practices smells like incompetence. We want certainty. For certainty, we are ready to prefer the professional illusionist to the empirically driven realist.

Second, we shouldn’t subject humans to experiments. Manipulating humans is rightly abhorred. We value freedom and self-fulfillment.

My hunch is that experimenting with better ways to work, will lead to more freedom and more self-fulfillment in the workplace. Why? The only way to get better decisions is to employ the abilities and senses of all the people in an organization. And we can’t get that level of engagement without offering more freedom and self-fulfillment.

The arch-capitalist quest for performance might just end up liberating people. 



  • Rosenzweig, Phil “The Halo Effect”
  • Bloom et al “Mangement Practices Across Firms and Countries”
  • More information about research on the link between management practices and companies performance can be found on


Posted by Frank Thun

Management. Systems. Liberation

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s