engineering-metrics

Track the four key metrics: deployment frequency, lead time for changes, change failure rate, and failed deployment recovery time. Gather 2-3 months of pipeline data, calculate your tier against DORA benchmarks, and focus improvement on the single weakest metric per quarter.

SPACE measures productivity across five dimensions: Satisfaction, Performance, Activity, Communication, and Efficiency. The core rule is to never use a single dimension alone — combine at least three to prevent gaming and get a balanced view of team health.

Use release pipeline run data for deployment frequency, AB# work item references to link commits through builds to releases for lead time, custom failed-deployment tags for change failure rate, and incident tracking integration for recovery time.

Velocity as a target triggers Goodhart's Law: teams inflate estimates, cherry-pick easy work, and sacrifice quality. Velocity is valid only as a team-internal sprint forecasting tool and is meaningless when compared across teams.

Cycle time measures active work start to completion, reflecting team efficiency. Lead time measures request or commit to completion, reflecting customer experience. Fast cycle time with slow lead time indicates a queue problem where work waits too long before starting.

Never use activity metrics as individual performance targets — setting commit count goals causes developers to make tiny commits. Use them only as directional aggregate signals, and never show individual-level activity data to leadership.