How can we measure user engagement time on Azure Data Explorer Web UI (dashboards and panels)?

Hi Saurav Datta,
This approach does not measure actual viewing or attention time, but it provides a reasonable approximation of when users were actively interacting with dashboards or panels in the **Azure Data Explorer Web UI.

As I mentioned earlier, Power BI is the best option for accurate engagement measurement. To achieve this, dashboards should be embedded in a custom application with client-side telemetry, or a platform with native engagement reporting like Power BI should be used.

Hope this helps. If you have any follow-up questions, please let me know. I would be happy to help.

There’s no built-in feature in Azure Data Explorer Web UI that directly reports “time spent” per user on dashboards or panels. However, the platform provides the building blocks to approximate engagement using Kusto logs and user analytics plugins.

A practical approach is:

1. Log dashboard access events
   • When dashboards are viewed via the Azure Data Explorer web UI, the underlying activity is KQL query execution against the cluster.
   • Use a telemetry/logging mechanism (for example, Application Insights or a custom log table in ADX) to capture at least:
     • User identifier
     • Dashboard/panel identifier
     • Timestamp of each view or interaction event
2. Model sessions and engagement windows in KQL
   • Once events are logged into a table in ADX, use Kusto’s user analytics plugins to derive engagement metrics.
   • For “how often” and “how many users” use dashboards (DAU/WAU/MAU-style engagement), use the activity_engagement plugin:

          Events
          | evaluate activity_engagement(
              "UserId",          // user identifier column
              "Timestamp",       // event time column
              1d,                 // inner window (e.g., daily activity)
              7d                  // outer window (e.g., weekly activity)
            )
     

     This calculates activity engagement ratios over sliding time windows and can be visualized in Azure Data Explorer dashboards.
3. Approximate “time spent” from event sequences
   • To estimate time spent per dashboard:
     • Treat each view/interaction as part of a session.
     • For each user and dashboard, sort events by time and compute the time difference to the next event or a max cap (for example, 10–15 minutes of inactivity ends the session).
     • Sum these durations per user and dashboard.
   • This logic is implemented with standard KQL (sessionization) on top of the logged events; the session_count plugin (listed in the user analytics plugins overview) can help count sessions, and custom KQL can extend that to approximate duration.
4. Visualize engagement in ADX dashboards
   • Use the Azure Data Explorer web UI query page to build KQL queries that:
     • Aggregate estimated time spent per user, per dashboard, per day/week.
     • Compute engagement ratios (DAU/WAU/MAU) with activity_engagement.
   • Export these queries as tiles into Azure Data Explorer dashboards for ongoing monitoring.

This approach uses ADX as both the storage and analytics engine for engagement telemetry: events are ingested into a table, KQL and user analytics plugins derive engagement metrics, and Azure Data Explorer dashboards visualize the results.

References:

• Visualize data with Azure Data Explorer dashboards
• Azure Data Explorer web UI query overview
• User analytics plugins
• activity_engagement plugin