Product Insights
Feature requests, bugs, and product feedback extracted automatically, each backed by the conversation that raised it.
Shown in Companion Dashboard only - this view isn't currently available on the org-wide Dashboard.
Product Insights automatically surfaces what people are asking for, complaining about, or suggesting - without you having to read through every conversation, and with a way to go read the real conversation whenever you need the full context.
What you see
A searchable, sortable, filterable list, one insight per row:
| Column | What it shows |
|---|---|
| Insight | The extracted piece of feedback. |
| Type | One of: feature request, pricing demand, usability, knowledge gap, bug, setup help, integration, or other. |
| Created | When this insight was first raised, in a conversation from the selected date range. |
| Updated | When this insight was most recently raised, in a conversation from the selected date range. |
| Count | How many conversations raised this same insight. |
Expand any row to see the quote pulled from the transcript that backs the insight up, plus a grid of every conversation it came from. Click a source card to open that conversation's full transcript in a new tab, or tap its info icon for a quick-look popover instead.
How to use this
- Share this list with your product team to inform roadmap decisions - recurring insights with a high count are the ones worth prioritizing.
- Filter by Type to focus on bugs when you need to triage urgently, versus feature requests when you're planning ahead.
- Expand a row before filing a ticket - the evidence quote and linked conversations give your team the exact context, in the user's own words, without you having to paraphrase it.
- Use alongside Top Keywords to see whether the topics people care about align with what they're explicitly requesting.
Insights are extracted automatically from conversation content for this companion. No setup required.
Last updated on
AI Insights
Sales opportunities, product feedback, failure patterns, and recommended actions - each linked back to the conversation it came from.
Recommended Actions
Suggested next steps to improve companion performance, generated from conversation patterns and linked back to the conversations behind each one.