Software and Chaotic Systems

In the book Sapiens, Yuval Noah Harari talks about two types of chaotic systems:

• Level 1 = does not react to predictions about it, e.g. the weather.
• Level 2 = reacts to predictions about it, e.g. markets, politics.

I propose that most software teams are a chaotic system as well. Traditional sales organizations require software teams to become more predictable — most often this is so they can sell features that haven’t been built yet. It may seem obvious, but this is a level 2 system, as the people on these teams will be impacted by their own predictions.

What happens to your team when you ask them to make a prediction? And, also ask them to commit to it? It’s worth considering if this it getting you more or less from your team.

As Harris points out, if a highly trusted super computer predicted the market price would go up to 100 dollars tomorrow, you likely will see that number reached — traders would react to the prediction. If a highly trusted software team predicted they would complete 100 points in two weeks, you likely will see that number reached. In both cases, you don’t know what could’ve been. Would you get more or less without the prediction?

Typically, high performing software teams (and all people) I’ve worked with will deliver more customer value when they are not asked to be predictable. It is as simple as what do you optimize for customer value or predictability — if you optimize for it you’ll get it; so, is predictability worth it? That’s up to you and your business. Personally,I’d rather have more customer value delivered.