What Gets Measured Gets Improved: Measuring Your Impact
Series: The Effective Engineer by Edmond Lau
Welcome back to my series on The Effective Engineer by Edmond Lau.
So far, we’ve covered how to prioritize high-impact work, leverage tools, and iterate quickly.
Now it’s time to talk about something that ties all those principles together: measurement.
Why Measuring Matters
You can’t improve what you don’t measure. Without clear metrics, you might spend weeks optimising something that doesn’t actually move the needle. Measurement provides clarity, it turns assumptions into evidence and helps you focus on what truly matters.
Metrics aren’t about micromanagement; they’re about alignment and learning. They help you understand whether your work is making the impact you intend.
What to Measure
Good metrics focus on outcomes, not outputs. They tell you whether your work is achieving its goal, not just that you’re busy.
Examples:
- Performance metrics: Latency, uptime, reliability, scalability.
- Productivity metrics: Deployment frequency, lead time for changes, defect rate.
- User metrics: Engagement, retention, satisfaction (e.g., NPS).
- Business metrics: Conversion rates, cost savings, revenue impact.
Before you start building, ask:
“What problem am I solving, and how will I know if I’ve solved it?”
Avoid Vanity Metrics
Vanity metrics look impressive but don’t reflect real progress. For example:
- Lines of code written
- Number of commits
- Story points completed
These measure activity, not impact. Focus on metrics that drive learning and decision making instead.
Tools for Measurement
Measurement should be automated and visible. Consider:
- Observability tools: Grafana, Datadog, New Relic.
- Analytics platforms: Amplitude, Mixpanel, Google Analytics.
- Dashboards: Real-time reporting on performance or business outcomes.
The goal is to make feedback loops tighter so data informs your next iteration faster.
My Experience
In one project, our team initially focused on velocity, how many features we could ship each sprint. But when we started measuring user adoption and satisfaction, we realised speed alone didn’t equal impact. Refocusing on metrics that captured real value led to better prioritisation and a much more effective engineering process.
Actionable Tips
- Define success early. Establish metrics before you start building.
- Focus on outcomes. Track user and business value, not just output.
- Automate measurement. Build dashboards or alerts for key KPIs.
- Review regularly. Use metrics to guide retrospectives and planning.
- Refine constantly. As goals change, evolve your metrics too.
Bringing It All Together
Measurement closes the loop in the effectiveness cycle. It connects prioritisation, iteration, and leverage, helping you make informed, high impact decisions. When you measure what matters, you stop guessing and start improving deliberately.
Next up: High Leverage Engineering - Multiplying Your Impact →