Process for people

Today, I attended a remote workshop on Defining, refining and aligning processes today, the third in a series of 7 workshops on engineering leadership.

I sure have heard or overheard people say that they hate process, and we generally favours less over more process at Automattic. Yet, what served us well when I joined in 2015 — when were a company of just under 300 — doesn’t always serve us well now that five years have passed and over 700 new colleagues have joined. I’ve found that many of the processes that we added in our team over the last three years were either very valuable — or easy enough to correct or abandon.

What is a good process for?

The first and obvious answer is that a good process helps solve a specific problem or achieve a specific outcome. But well-defined and -documented processes can also make a group more inclusive. One of the panelists mentioned a feminist essay from Jo Freeman, written in the 1970s:

From this essay:

As long as the structure of the group is informal, the rules of how decisions are made are known only to a few and awareness of power is limited to those who know the rules.

Jo Freeman, The tyranny of structurelessness

Applied to our field, this may mean that a clear process for escalating and addressing issues can enable junior developers to become more productive and to progress to become senior developers sooner. Documenting clear expectations beforehand can make performance reviews fairer and helps newcomers navigate their new environment more successfully. It removes the burden to gather information or clues from long-time or successful colleagues by spelling the rules out clearly. Processes that come from a perspective of caring can make a group more efficient and inclusive. Good processes are processes for people.

What defines a good process?

Adrienne highlighted three characteristics of a good process in her talk What makes a good process?:

It is created by & for the people who use it.

It is documented & changeable.

It has reflection & evaluation built-in.

I think her three points are a very good guideline and everything I can think of wanting to add is already included in them.


By Sirin Odrowski

Data Science and Engineering Manager at Automattic

%d bloggers like this: