How We Measure Engineering Productivity

Most engineering leaders know when something is wrong. Delivery is slowing down, morale is dipping, incidents are increasing. But knowing something is wrong and knowing exactly what to fix are very different things. That's where measurement comes in.

At Arc&Delta, we measure engineering productivity across four dimensions: speed, effectiveness, quality, and business impact. This draws from the DX Core 4 framework — the same approach used by engineering organisations at Meta, Microsoft, and Uber — combined with DORA metrics that have become the industry standard for delivery performance.

The Four Dimensions

Speed measures how quickly work moves through your system. The key metrics are deployment frequency, lead time from first commit to production, and perceived rate of delivery. A team that deploys frequently with short lead times is a team with a healthy, low-friction delivery system.

Effectiveness measures whether your engineers can actually do their best work. This includes the Developer Experience Index — a measure of how much friction engineers encounter day to day — alongside ease of delivery and regrettable attrition. If your best engineers are leaving, something in the system is broken.

Quality measures the stability of what you ship. Change failure rate and mean time to recovery are the core signals here. A high change failure rate tells you your testing, review, or deployment processes need attention. A long recovery time tells you your incident response and system design need work.

Business impact measures whether engineering effort is translating into outcomes. What percentage of time is spent on new capabilities versus maintenance? Is R&D spend generating return? These questions connect engineering performance to the metrics that matter in the boardroom.

Why Most Teams Don't Measure This Way

The most common mistake we see is measuring outputs instead of outcomes. Ticket velocity, lines of code, and utilisation rates feel like productivity metrics — but they measure activity, not impact. A team closing 50 tickets a week while the system becomes harder to maintain is not a high-performing team.

The second most common mistake is measuring in isolation. Speed metrics without quality metrics create pressure to ship fast and break things. Quality metrics without speed metrics create overly cautious teams that ship too slowly. The four dimensions only work when measured together.

How We Apply This at Arc&Delta

Most organisations we work with don't have these metrics in place when we arrive. That's not a problem — it's often part of why they need us.

The first thing we do is establish visibility. Before we recommend anything, we work with your team to understand where you actually are — through a combination of structured interviews, delivery data, and direct observation. That baseline, however informal at first, becomes the reference point for every decision we make.

As the engagement progresses, we introduce the right measurement practices for your stage and context. Not every company needs a full DORA dashboard on day one. But every company benefits from knowing whether delivery is getting faster, whether quality is improving, and whether engineering time is being spent on the right things.

By the end of an engagement, you have both the outcomes and the visibility to sustain them. That's what separates a transformation that sticks from one that fades the moment we leave.