Even with thorough documentation, there are always topics your users need that you have not covered yet. Userorbit's autopilot feature analyzes your help center content against real user behavior to identify these gaps and suggest where to focus your writing effort next.
What autopilot does
Autopilot is an automated analysis tool that examines several data sources to find documentation gaps:
- AI conversation logs — Questions users asked that the AI could not fully answer, indicating missing or incomplete content.
- Search queries — Terms users searched for in your help center that returned no results or low-relevance results.
- Escalation patterns — Topics where AI conversations frequently escalate to human agents, suggesting the documentation is insufficient.
- Article coverage — A comparison of your product features and workflows against the topics your articles actually cover.
The result is a prioritized list of content suggestions, ranked by how much impact they would have on reducing support load and improving the AI conversation experience.
How to access autopilot
To view your autopilot analysis:
- Navigate to the Help Center section in Userorbit.
- Click the Autopilot tab or look for the coverage analysis section in your help center dashboard.
- Userorbit displays the latest analysis results, including suggested topics, gap severity, and estimated impact.
Autopilot runs automatically on a regular schedule, so results stay current as your content and user behavior evolve. You can also trigger a manual refresh if you have recently published several new articles and want to see updated results.
Interpreting the results
Autopilot presents its findings in several ways:
Gap severity
Each identified gap is assigned a severity level based on how frequently users encounter it and how poorly the current content addresses it:
- High — Many users are asking about this topic and getting no useful answer. Prioritize writing this content.
- Medium — Some users hit this gap, or existing content partially covers it but needs expansion.
- Low — A small number of queries relate to this topic, or existing content is close but could be improved.
Suggested topics
For each gap, autopilot suggests a topic title and brief description of what the article should cover. These suggestions are based on the actual language users use in their questions and searches, so they reflect real needs rather than assumptions.
Related existing articles
Autopilot also identifies existing articles that are close to covering the gap. Sometimes the fix is updating an existing article rather than writing a new one — adding a missing section, clarifying a confusing explanation, or covering an edge case.
Acting on suggestions
Once you have reviewed the autopilot results, take action:
- Prioritize by impact — Start with high-severity gaps that affect the most users.
- Check for quick wins — Some gaps can be closed by adding a paragraph to an existing article rather than writing an entirely new one.
- Create new articles — For topics that are not covered at all, create a new article using the suggested topic as a starting point.
- Review after publishing — After publishing new content, check autopilot again in a few days to see if the gap has been resolved.
Building a continuous improvement loop
Autopilot is most valuable when used as part of a regular workflow rather than a one-time audit. Consider reviewing autopilot results weekly or biweekly as part of your content planning. Over time, you will see the number of high-severity gaps decrease as your documentation matures, and the AI conversation success rate will improve alongside it.
By letting real user behavior guide your documentation priorities, you ensure that every article you write has maximum impact on the user experience.