Skip to content

The Analytics dashboard gives tenant administrators a comprehensive view of how your organization uses the GrndCtrl AI chat. Track usage trends over time, see which offices and users are most active, analyze query categories, monitor feedback rates, and drill into individual conversations — all from one interactive dashboard.

Click the Apps button — the grid icon in the header, available on every page — and choose Conversation Analytics. See Finding and Opening Your Apps if it is not listed there.

The dashboard loads with a progress overlay showing its loading stages.

The Conversation Analytics dashboard, showing KPI cards above the conversation data table

At the top of the dashboard, set the date range for the data:

  • Start Date — defaults to the product launch date (September 2025)
  • End Date — leave empty for “up to today” or pick a specific date
  • Click Apply Date Range to reload the data
  • Click the Refresh icon next to the conversation count to reload without changing dates

Six KPI cards appear at the top:

Card What It Shows
Unique Users Count of distinct users who chatted in the date range
Total Sessions Number of conversation sessions started
Total Requests Total user messages sent to the AI
Avg Daily Requests Average messages per day
Avg Weekly Requests Average messages per week
Avg Monthly Requests Average messages per month

Below the usage KPIs, four feedback cards show response quality metrics:

Card What It Shows
Rated Messages Total messages that received thumbs up or down (clickable — filters the table to only rated messages)
Positive Feedback (green) Count of thumbs-up ratings
Negative Feedback (red) Count of thumbs-down ratings
Satisfaction Rate Percentage: positive / rated messages

Click View Charts to expand the charts section (click Hide Charts to collapse it). All charts are interactive — click any data point to cross-filter the conversation table below.

A time series chart showing conversation volume. It supports three levels of drill-down:

  • Monthly — click a month to drill into weekly view
  • Weekly — click a week to drill into daily view
  • Daily — the most granular level

A breadcrumb trail at the top shows your current drill level, and a Back button returns to the previous level.

Horizontal bar chart showing requests per office and unique users per office. Click any bar to filter the table to that office.

Horizontal bar chart showing the top 8 users by request count. Click a bar to filter the table to that user.

Bar chart showing request volume for each hour (0–23). Click a bar to filter the table to messages sent during that hour.

Pie chart showing the distribution of message classifications (query types). Click any slice to filter the table to that category.

When you click a chart element, it acts as a filter. Active filters appear as blue chips above the table. You can:

  • Click the X on any chip to remove that filter
  • Click Clear all to reset all filters
  • Stack multiple filters (e.g., filter by office AND hour)

Below the charts, a sortable data table lists individual messages:

Column Description
Seq Sequence number (chronological, oldest = 1)
Timestamp Date and time of message
User Who sent the message
Office User’s office location
Category Classification badge for the query type
Rating Thumbs up, thumbs down, or dash (no rating)
Prompt User’s message (truncated — hover for full text)
Response AI’s response (truncated — hover for full text)

Click any column header to sort. Click a row to open the Conversation Detail modal with the full message content.

Solutions:

  • Expand the date range — the start date defaults to September 2025, but your data may be more recent
  • Click Apply Date Range after changing dates
  • Check that users in your tenant have been chatting with the AI

Solutions:

  • Large date ranges with thousands of messages take longer to process
  • Narrow the date range to improve load times
  • The progress overlay shows the current loading stage so you know it’s working