Spotter Conversations Liveboard

Use the Spotter Conversations Liveboard to get insight on how your users are interacting with Spotter. You can gain insight into user engagement and adoption metrics for Spotter, track query trends and understand user intent, monitor follow-up actions (including query refinements and edits within conversations), and analyze feedback to identify training gaps and improve dataset coverage.

Accessing the Liveboard

You must have admin privileges to view this Liveboard.

To view this Liveboard, navigate to the Insights tab, select Liveboards from the left menu, and search for Spotter Conversations.

Spotter Conversations Liveboard

Navigating the Liveboard

You can filter the Liveboard by timestamp, Org, Worksheet, conversation origin, and user name using the filters at the top of the Liveboard.

Understanding the metrics

This Liveboard provides insight into the following key areas:

Overview Metrics

Spotter conversations overview

  • Active users (weekly): Shows the number of users actively engaging with Spotter.

  • Conversations (weekly): Tracks the total number of Spotter conversations that users had.

  • Average conversation length (weekly): Provides the average number of questions asked in a conversation.

  • Conversations with feedback (weekly): Indicates the number of conversations where users either downvoted or upvoted a response.

  • Most active users: Identifies users who have had the highest number of conversations on Spotter.

  • Most active Worksheets: Highlights which Worksheets are used most frequently.

  • Conversation length distribution: Displays a frequency distribution of the number of questions asked in a conversation.

  • Feedback participation: Measures how often users provide feedback by showing the percentage of conversations with upvotes or downvotes. Since a conversation can have both, the total may exceed the number of unique conversations.

Conversations Analysis

Spotter Conversations page

  • Total conversations: Overall volume of Spotter interactions.

  • Downvoted conversations: Total number of conversations where users downvoted a response, indicating a need for coaching.

  • Conversations with downvoted responses: Displays a log of user interactions in downvoted conversations, sorted by timestamp. The most recent conversations appear at the top.

  • Complete conversations log: Shows every user interaction as a separate row, sorted by timestamp, with the most recent interactions appearing first.

Best practices for using the Liveboard

Follow this structured workflow to analyze data and take action:

  1. Start with overview metrics.

    • Review active users, conversation trends, and feedback volume to understand general engagement.

  2. Identify issues and patterns.

    • Look at downvoted responses and refinements to identify problematic areas.

  3. Take action.

    • Adjust Spotter coaching, refine datasets, or offer user training and guidance to improve AI performance.

  4. Monitor long-term impact.

    • Use trend analysis to measure improvements in AI accuracy and adoption over time.

Was this page helpful?
×