The Mirage of Measurement
Engineering leaders rarely set out to measure the wrong things. Metrics begin as tools for coordination. They are intended as a way to make progress visible and predictable across complex, distributed teams. They exist to reduce uncertainty, to give shape to effort, and to help align decision-makers who can't see the work directly.
But visibility can ofter be a double-edged sword. The same metrics meant to clarify progress soon become artefacts for presentation. They become numbers built not for understanding, but for defensibility. Many leaders would lead better if they didn't have to translate reality into slides for executives.
The dashboards look healthy. Story points rise, incidents fall, OKRs tick upward. Yet product momentum slows. Teams are moving faster but not necessarily in the right direction. It's a quiet, familiar paradox, and presents the illusion of control achieved through the measurement of motion.
How Metrics Drift from Meaning
The moment a metric becomes a target, it begins to distort behaviour. Story sizing is trimmed for efficiency. Reliability numbers improve because incidents are reclassified. Customer satisfaction rises because surveys are redesigned. Progress becomes a choreography of movement designed to look like advancement.
Leaders rarely notice when this drift begins, because metrics are seductive. They promise clarity and accountability. But what they really deliver is the comfort of precision without the burden of understanding.
And precision is not the same as accuracy. Precision gives the illusion of control. The numbers that look neat, the trends that appear stable, and graphs that provide comfort anxious executives. But accuracy is different because it demands context, curiosity, and confronts, rather than hides, uncertainty. Precision describes how exact a measurement is; accuracy tells you whether it's true. Many teams master the former while losing the latter.
The team optimises for the number and now the number no longer represents reality. Metrics aren't the enemy. The enemy is the disconnection between what is being measured and what actually matters — a disconnection widened every time we measure to reassure others instead of to learn ourselves.
A Simple Example
One product team I observed measured success by "stories completed per sprint." It made progress look sharp and consistent — until quality complaints began to rise. Engineers had quietly adapted: they sliced stories thinner, avoided risky improvements, and deferred hard work to keep the charts green.
The fix wasn't radical. Leadership reframed the metric from "stories completed" to "stories validated in production." Suddenly, speed alone no longer signalled success. Engineers stopped optimising for closure and started optimising for outcome. They began releasing smaller experiments, collecting real-world feedback, and improving faster than before.
The metric didn't get better because it was bigger; it got better because it was closer to purpose.
The Missing Layer: Guiding Policies
Most organisations move from mission straight to action — "We want to deliver faster, so we will measure velocity." But the connective tissue between those two layers is missing: the guiding policies that translate purpose into principle.
Guiding policies answer the quiet questions that metrics never can:
- What does "improvement" mean in this context?
- Where is efficiency an advantage, and where does it invite fragility?
- What trade-offs are acceptable?
Without those policies, actions are untethered from intent. Teams work harder but not necessarily wiser. And metrics — deprived of a philosophical anchor — evolve into governance theatre.
Restoring Causality
The solution is not to abandon metrics, but to rebuild their hierarchy. Start with the mission. From there, articulate guiding policies — the heuristics that express how you intend to pursue it. Only then decide which actions and measurements serve that framework.
When metrics arise naturally from policy, they regain their meaning. They stop dictating behaviour and start reflecting it. Teams understand why they are measuring, not just what. Leaders interpret numbers as signals, not scores.
This restores causality: mission → policy → action → metric. When that sequence is reversed, organisations confuse movement for momentum and data for direction.
The Real Measure of Progress
Engineering excellence isn't about going faster. It's about knowing what not to accelerate.
Metrics are useful only when they illuminate the path to meaning. When they become an end in themselves, they erode judgment, flatten creativity, and reward the wrong instincts.
The most mature teams are not those that produce the most data, but those that can explain — clearly and coherently — why they measure what they measure.
Metrics are shadows of reality — projections of something more complex and alive. They show its outline, not its substance. When organisations start optimising the shadow, the source — purpose, curiosity, and judgment — begins to fade.