Marcus Teo · 2025-11-04

Release board metrics that are not vanity theatre

Dashboards love big integers. Testing leaders feel pressure to show upward trends even when quality is flat. In this field note we document four charts our alumni keep after graduating from the Release readiness track.

The first tracks mean time to reproduce a customer-reported defect with a linked charter. The second compares contract test failures against production incidents thirty days later. Neither chart is glamorous; both survive skeptical finance reviews because they map to dollars saved or lost.

We also explain how to retire metrics that incentivise the wrong behaviour—classic example: raw automated case counts that reward duplication instead of meaningful assertions.

Use the reflection prompts at the end with your release manager; they are written so you can paste them into Confluence without sounding like a manifesto.