Use the Adoption Dashboard to ensure AI Sidekick is giving accurate, helpful answers backed by trustworthy content.
Monitoring AI Sidekick Effectiveness
Start here when: You want to ensure AI Sidekick is giving accurate, helpful answers and that the content powering it is trustworthy.
Step 1: Review Recent Sidekick Chats
The Recent Sidekick Chats table is your most detailed view of AI Sidekick performance. For each interaction you can see the question asked, the response given, the source content used, and whether the user provided feedback.
Step 2: Prioritize Negative Feedback
Filter the Recent Sidekick Chats table to surface interactions with negative feedback first. For each one:
- Read the question and response in full
- Check whether the source content is accurate and up to date
- Update the source content immediately if needed
Step 3: Monitor Source Quality
Review the Content Highlights table to see which content AI Sidekick is referencing most frequently. High AI response counts on a piece of content mean it has significant influence on what users learn — it must be accurate and current.
Best Practices
- Review Recent Sidekick Chats weekly — negative feedback is a direct signal that source content needs attention, and most users won't leave feedback, so proactive monitoring matters.
- Cross-reference Content Highlights with Last Edited Date in the Content Engagement table — frequently sourced but stale content is a high risk.
- Set an alert on Recent Sidekick Chats for "Results have changed" so you're notified when new negative feedback comes in, and don't rely on AI themes alone — individual chat reviews reveal the real issues.