Skip to content

System Signal Tracking (DORA)

Lead time, change failure rate, deploy frequency, time to restore. The health of the delivery machinery itself — independent of any one feature. A team can ship one good feature on a dying chain. DORA catches that.

Owners: Tech Lead, Leadership Phase it lives in: After We Build (Volume V) The corpus principle this enacts: The portfolio view is a diagnostic tool — DORA diagnoses the machinery; chain-aware root causes diagnose the meaning.

Where it lives in the chain

The four metrics

MetricReads as
Lead time (commit → production)How thick the chain is. Long lead time means a bottleneck somewhere — review, CI, manual gates.
Change failure rateHow often the chain ships something that has to be rolled back. Catches insufficient testing, missing flags, brittle pipeline.
Deploy frequencyHow often the chain is willing to commit. Low frequency hides problems — large changes that ship rarely accumulate risk.
Time to restore (MTTR)How quickly the chain recovers when it breaks. Catches runbook freshness, incident discipline, on-call training.

How to do this

  • Track at portfolio level. Per product, per quarter. Per-story DORA is noise.
  • Watch the trend, not the absolute. A team going from 1 deploy/week to 3 deploys/week is a story; a team at 5 deploys/week steady is in maintenance mode.
  • Read DORA alongside chain-aware root causes. DORA says how well the machine runs; root causes say how well the chain carries meaning. Both matter. Neither alone is sufficient.

What good practice looks like

DORA is part of the quarterly portfolio review, alongside VRI and chain-level distribution. The team reads:

  • Lead time trending up — means review is slow, pipeline is brittle, or branches are long. Investigate which.
  • Change failure rate above 15% — means the chain of catches is missing a layer. Audit the seven stages.
  • Deploy frequency below 1/week — means the team is batching. Ask whether trunk-based discipline is intact.
  • MTTR climbing — means runbooks are decaying, the on-call is undertrained, or the system has grown harder to debug.

DORA is not a team scorecard. It is a chain diagnostic. A team measured on DORA gamed it; a team that watches DORA together fixes the chain.

200apps · How We Work · NWIRE