The software industry and an unhealthy obsession with metrics

We can’t have one OKR. We can’t have one KPI. We can’t have one goal. We can’t have one focus for the quarter. Why are we slow to deliver new features to our customers?

Scope grows quickly. Focus diminishes rapidly. It all starts before you even consider what the solution will be.

Senior leader: “If you have that OKR, then you should also do something about this related thing”
You: “Sure, what order should we do it in? What’s more important?”
Senior leader: “We need both at the same time, don’t ask me to do my job and prioritise!”
You: “Sigh. Ok.”

The quarter goes by, you kinda do a bit of everything, but it feels like you haven’t really done anything. If you’re lucky you will have a candid retro about the need for focus, people will agree and say all the right things, promising next quarter will be different.

Next quarter comes around and it’s much of the same. You spend a month planning. Your single OKR become 2 or 3. Your focus thins out, the team is stretched and we repeat the corporate cycle.

There is an unhealthy obsession with measuring and defining metrics in software organisations. This is all under the guise of being “data driven” but as usual, anything good is taken to the extreme (like test driven development, and aiming for 100% test coverage…)

Critical thinking and nuance be damned, we’re going to make sure we have data before making a decision. So let’s setup our tracking and visualisations before getting started. Since we defined so many OKRs, KPIs and goals we need to be able to track them. What’s that? It costs time and effort to set that all up? That time bites into actually doing the work? Ah that’s ok, we can do it all! We’re a high performing organisation! (then why do we need the KPIs?)