[Reading time: 2 minutes 21 seconds]
Bullshit.
“The law of conservation of bullshit”
Call me childish, but it’s all I can do not to cackle with glee at that expression.
It’s kind of what’s behind many rules or observations about organisation and team size.
If you’re just a handful of engineers, sharing information is not so difficult. Everybody gets together and communicates naturally. Information flows as though on its own. Everybody knows everything that’s going on.
As a team grows, however, this becomes more and more cumbersome.
Meeting invites become more frequent and less interesting. Endless daily standups become an exercise in restraint.
Eventually, the engineers collectively get sick of the bullshit, and a new organisation is instituted: you split the gang up into multiple teams with distinct responsibilities – and communication paths.
A separate structure is created on top of these several teams: maybe team leads or some other function. They get together and share information across teams.
This works fine for a while, until… the teams become so numerous that bullshit starts to get out of hand as far as this management group as well.
So the law of conservation of bullshit kicks in again, shaping a desire to create even more insulation between teams.
The question is: what do you do then?
You can’t keep building more indirect reporting/alignment systems (well, yes, you can, but it won’t be pretty).
So as you grow, you’ll need to find other ways to cut down on the bullshit.
One such way is… affording your people more autonomy.
Autonomy.
In other words: make them be efficient and effective at doing their jobs, while lessening the need to communicate.
Autonomy is one of these single words you can write entire books about though: many things need to align for autonomy to arise.
First of all, you need to have an appropriate culture – for example, one where autonomy isn’t eyed with suspicion, but encouraged.
You need to have an architecture that exhibits appropriate separation of concerns, so that teams can build on what their colleagues did without asking for help or permission (staying in conversation is still a good idea though).
You need to have the team structure that fits the architecture, and likewise enables self-sufficiency in teams (e.g. by providing the required skills).
And you need to have processes in place which can deal with things happening autonomously.
Getting there can be a lot of work, but you’ll find it’s worth it: this way, you can get a lot of stuff done, without much of the bullshit.
Isn’t that nice?