Analytics
The Analytics section gives you a clear view of how people are using your portal, how well the AI assistant is performing, and where your content may have gaps. You will find it under the Analytics tab within each portal's admin console.
Overview Dashboard
The Overview tab is your at-a-glance summary of portal activity.
Key Metrics
Total Queries
The total number of questions sent to the AI assistant across all time
Active Users (30 days)
How many distinct visitors interacted with the portal in the past month
Satisfaction Rate
The proportion of AI responses that users rated positively (thumbs up)
Coverage
What percentage of your product catalog is currently published and live
Escalations
How many conversations included a request for human support
Data Sources
The total number of data sources included
Most Asked Questions
A ranked list of the ten questions asked most frequently across all portal sessions.
This is very useful for identifying content gaps, if users keep asking about something and the AI is struggling to answer well, it may signal that you need richer documentation for those topics.
Top Searched Products
A ranked list of the ten products most frequently discussed or searched.
Use this to prioritize content quality, if a product is getting a lot of attention in the portal, make sure its catalog entry is complete and up to date.
Users
The Users tab shows a breakdown of engagement by individual portal visitor.
For each user, you can see:
User
Email address or anonymous visitor identifier
Sessions
How many chat sessions they have started
Messages
Total messages they have sent to the assistant
Products Explored
How many distinct products they have viewed or discussed
Last Active
When they most recently used the portal
Feedback
How many thumbs up and thumbs down ratings they have submitted
Click any row to view that user's full session history and activity timeline.
Feedback
The Feedback tab shows a log of all ratings that portal visitors have submitted on AI responses, both positive (thumbs up) and negative (thumbs down).
For each feedback entry you can see:
Who submitted it and when
Whether it was positive or negative
The AI response that was rated
The question or message that prompted that response
A link to the full session for context
Filtering Feedback
You can filter the feedback log by: All Feedback, Positive, Negative.
Reviewing negative feedback regularly is one of the most effective ways to identify where the AI is falling short and what content improvements will have the biggest impact.
Last updated