Skip to main content

Common Questions

The Common Questions tab shows which questions are asked most frequently of your Kapa instance. The platform automatically clusters similar questions to help you identify patterns and themes in user inquiries.

Common Questions screen

Clustering methodology​

Kapa uses semantic clustering to group similar questions together, regardless of their exact wording. This helps you understand the broader topics users are interested in, rather than focusing on individual questions.

Questions are clustered on a weekly, monthly, and quarterly basis. You can switch the interval period in the UI and use the arrow navigation to go backwards and forwards in time.

Certainty filter​

Questions can be filtered based on whether Kapa was able to provide a certain answer or not. The platform clusters all uncertain questions separately, allowing you to identify areas where Kapa couldn't find relevant information.

The uncertain questions represent areas where Kapa couldn't find content in your documentation to reliably answer the question. Reviewing these clusters can help identify potential documentation gaps.

Editing clusters​

The clustering is fully automated, but you can refine it manually to better suit your needs:

  • Delete clusters: Remove clusters you find irrelevant or don't agree with
  • Rename clusters: Change the automatically generated titles to be more descriptive or accurate

FAQ​

What time horizons can I look at?​

Questions are clustered on a weekly, monthly, and quarterly basis. You can switch the interval period in the UI and use the arrow navigation to go backwards and forwards in time to analyze how patterns change over time.

Why do I not see any clusters?​

Your Kapa instance needs to have had at least 150 separate conversations in a time period before the platform starts generating clusters. This minimum threshold ensures that the semantic clustering is meaningful.