The Zeno effect in business. Watching stops development

The first book coming out pushed me to dust off an idea I have been turning over for about three years. I am writing a second book. Not about business. It will be a synthesis of quantum mechanics, genetics, evolution, and psychology. And one night, digging through quantum mechanics, I hit the Zeno effect.

The physics is simple. If you watch an unstable particle continuously, it stops evolving. Observation freezes its state. It does not slow the particle. It freezes it. Recorded in 1977, confirmed many times since. Pure physics.

I put the laptop down and stared at the ceiling. Because I had just understood: this is not only about particles.

Observation freezes

When you start controlling people hard — daily reports, constant check-ins, a KPI for every sneeze — you get exactly what the physics describes. People freeze in the state you measure. They stop doing anything except the thing you control them for.

This is not betrayal and it is not laziness. It is a rational adaptation to the environment. A person starts optimizing the metric, not the result. Not because they are a bad employee. Because the system says, clearly, this is where we are looking. This is what we will ask about.

Daniel Pink, in Drive: The Surprising Truth About What Motivates Us (a book on the nature of motivation, built on 50 years of psychological research, 2009), showed this through a broad reading of the science of motivation. His conclusion is hard. Sticks and carrots work only where the task is mechanical and does not require thinking. For everything else — for making, searching, creating something new — outside control is not merely useless. It lowers the result. MIT researchers, in a series of experiments, found that the higher the money offered for a task that takes cognitive effort, the worse people do it. The reward shifts attention from the task to the outside stimulus, and thinking narrows at the moment it should widen.

Observation works the same way. A person under observation thinks not about the task, but about how their work looks to the observer.

The effect has two sides

Here is what I understood from my own experience of managing. Observation freezes a state. That is not always bad. It depends on one question: which state do you want to fix in place?

If the function has already been found, tuned, and it works, and you need stability and predictability, then watch. That is what Frederick Taylor was doing at the start of the twentieth century when he built scientific management. He studied the work, found the best way to do the task, standardized it, and then watched the execution. Ford, McDonald’s, and Amazon in its operations logistics work on the same principle. Where you need predictability and as little variation as possible, frequent observation is not control for its own sake. It is an architectural decision.

Yes, people with creative ambitions leave those functions. That is normal. It is the price of the choice. The job of such a function is to run like a clock, not to look for unusual paths.

Observation freezes a state. If the state is the one you want, freeze it. If it is not yet, step back and let the system evolve.

Where you need evolution, observation kills it

A new product is a different matter. Searching for a niche. Creative decisions. Anything that means getting through uncertainty. There, hard observation kills the very thing you hired people for.

People start fitting themselves to the metrics and the deadlines, and they hand in not the best solution but the one that can be handed in on time. Speed replaces quality. Controllability replaces search.

Google understood this earlier than many. The “20% time” policy — engineers get every fifth working week with no tasks and no reporting — produced Gmail, Google Maps, and Google News. Not because the company had time to spare. Because between the points of observation the system had room to evolve.

Pink calls this autonomy, one of the three core elements of intrinsic motivation, along with mastery and purpose. On his data, autonomy correlates most strongly with how innovative a team is. And it is the first thing micromanagement destroys.

How to use this

Before you build a control system, answer one question. What state does this function need: a stable one, or one that is still developing?

If it should be stable, watch often. Standardize. Measure. Operations, a call center, logistics, the accounts. Predictability is needed there, and frequent observation provides it. That is Taylor, and in his own field he is right.

If it should be developing, create gaps with no observation. Give people space between the control points. Product, R&D, strategy, new markets. People there have to have time to try, to be wrong, and to try again before anyone comes with questions. Otherwise they will report on attempts instead of making them.

Physicists solved it like this: if you measure less often, the particle has time to evolve between observations.

The same principle. Only it is not about particles.