Skip to content

Example: a documentation review workflow

View as Markdown

A new feature ships. The writer has drafted a new task article documenting it, and it needs to go out with the release.

  1. Self-review first. The writer runs the documentation quality checklist against their own draft before asking anyone else to look at it - catching the obvious issues here is cheaper than a reviewer catching them.

  2. Technical review. The writer sends a review request to the engineer who built the feature, scoped explicitly: “Confirm the steps still match the shipped UI - I’m not asking about wording.” The engineer works through the technical review checklist and leaves blocking comments only on genuine inaccuracies.

  3. Author addresses feedback. Every blocking comment gets a fix or a reply explaining why it doesn’t apply. Non-blocking suggestions are triaged - some taken, some explicitly declined with a reason.

  4. Editorial pass. A second reviewer - or the same writer, after time away from the draft - checks structure, tone, and consistency with the rest of the site. This pass explicitly does not re-litigate technical accuracy; that was step 2’s job.

  5. Release-readiness check. Before the feature ships, the release readiness checklist confirms the new article is linked from navigation, builds cleanly, and (if applicable) has a translation plan.

  6. Publish alongside the feature, not after it - see integrating documentation into development.

Technical review and editorial review ask fundamentally different questions - “is this true?” versus “is this clear?” - and asking one reviewer to hold both hats at once tends to produce a review that does neither well. Splitting the passes also means feedback doesn’t collide: an engineer’s technical correction won’t get lost among a dozen wording nitpicks, and vice versa.