Why Proxy Metrics Can Drive Continuous Improvement
Subtitle: How Don Reinertsen’s idea helps teams focus on what really drives improvement — not just what is measured.
The Problem with Traditional Metrics
In the past I have observed a recurring issue: metric collection and reporting become detached from actual improvement.
Teams gather numbers. Dashboards fill with charts. Reports circulate every week. Yet when you look closer, real outcomes related improvement don’t move much. Worse still, nobody understands why things change.
Why? Because many of the metrics being tracked aren’t connected to the system’s performance drivers. Instead, they’re by-products of activity, not indicators of progress.
Too Much Focus on Quality Exercises — Not Enough on Results
Many companies spend huge amounts of time and effort on traditional quality exercises.
These activities create “busy work”, but they often miss the connection to actual outcomes. Quality is treated as an event a monthly review, a meeting, a report, rather than as a continuous flow of improvement actions.
Without understanding how quality relates to improvement, teams risk and often ending up “optimising” what is visible.
The Car Dashboard Analogy
Think about driving a car.
A modern vehicle collects thousands of measurements every second — from engine temperature to wheel rotation, airflow, and fuel mix. Yet on your dashboard, you see only a handful of key signals: speed, fuel level, engine warning, and maybe navigation guidance.
That’s because drivers need information that helps them act, not raw data.
The same principle applies to organizations. We don’t need to see everything that’s measurable. We need to see the few signals that truly guide decisions and behaviour.
Introducing Proxy Metrics
Don Reinertsen, in The Principles of Product Development Flow, introduced the idea of proxy metrics — measurable indicators that stand in for deeper, harder-to-measure system qualities.
A proxy metric doesn’t try to capture the whole system. Instead, it represents a reliable signal that correlates with what we really care about.
By focusing on the proxies organizations can change the lower level concrete metrics and trust that the Business metrics will in turn improve.
Connecting Proxy Metrics with Continuous Delivery
In their book Continuous Delivery, Jez Humble and David Farley describe a culture of constant feedback, fast flow, and automation.
Reinertsen’s proxy metric concept fits perfectly here.
Continuous Delivery thrives when teams can focus on the proxy. Proxy metrics make that possible — especially when outcomes like customer satisfaction or market response take weeks or months to appear.
By combining Reinertsen’s thinking with Continuous Delivery principles, I promote 3 proxy metrics to build organizations that are fast, focused, and trustworthy.
How to Apply This Thinking
Start simple.
- Identify your required outcomes. What really defines success — customer retention, deployment frequency, decision speed?
- Decide of a minimum set of metrics that align with business outcomes. Which ones inspect and report.
- Find proxy candidates. Decide on the underlying proxies that when modified move the needle on your business performance.
- Rinse and Repeat. Keep only the signals that support decisions.
Your goal isn’t more measurement — it’s better correlation between what you measure and what you manage.
Why It Matters
Every system produces more data than we can handle. The real advantage comes from knowing which measurements help us steer, not which fill our reports.
Proxy metrics help organizations stay agile, align effort with impact, and create a shared language between strategy and execution.
They’re the bridge between what we can see and what truly matters.
If you want to know more about proxy metrics and how they can improve your business performance. Fill in the contact form below.