Why You Should Measure Productivity

Problem

There are a lot of reasons that can feed into whether a team is productive or not. One of the bigger barriers I see is that most engineering leaders think that understanding and evaluating productivity is bad.

Actions taken

Productivity in Engineering has become a bit of a bad word and even measuring it can be frowned upon. However, if you don't measure productivity you won't have as much ability to improve, and as a leader, one of your main jobs is to help your team to be as productive as they can be. Productivity can be a really helpful indicator. Always start from a team or project level and expect that whatever metric you come up with will be, at best, directional. The point of a metric is to give an indication of when you should raise a conversation about what could be done better to help productivity. The specific metrics I find most useful are volume metrics (e.g. stories, story points, tickets) and how often the team hits milestones they set. These metrics help you to detect anomalies. For example, if the number of tickets you close dips for a couple of weeks you can then investigate what is going on. Usually, there's an obvious reason for these types of anomalies. If there isn't an obvious reason, that usually provides you with an opportunity to improve something - e.g. your tooling may be breaking down or the number of distractions may have gone up.

Lessons learned

A lot of the stigma around measuring productivity comes from people who don't understand engineering asking for arbitrary metrics and from unskilled leaders not knowing how to use the information and using it to penalize teams. Metrics of productivity are really useful. However, when using them, just keep in mind that metrics don't tell you the story, they just tell you that there's a story there.