Turn customer feedback
into patterns you can act on.
Use AI to analyze collected feedback, group similar requests and find recurring problems. Keep the original customer context, review the conclusions and connect them to your backlog.
Analyze feedback with AIFind the recurring problem behind different requests
Acme asks for invoice CSV export through a widget, then repeats the need in Slack. Birch asks for the same outcome in Discord. Delta asks for weekly invoice emails. The assistant can propose themes; your team checks whether the requests actually describe the same need.
- Select the evidence: choose the requests, reviews or supplied excerpts and the period to analyze.
- Group similar requests: ask for proposed themes with links to the original feedback.
- Find repeated problems: separate repeated mentions, distinct customers and different requirements.
- Review: check the source wording, correct assumptions and confirm possible duplicates.
- Connect the backlog: link the reviewed Idea to existing work, or prepare a new task when the need is distinct.
- Decide with the team: use the evidence alongside impact, effort and customer context.
┌─ EVIDENCE FOR TEAM PRIORITIES ───┐ │ Review impact and effort │ │ Add known customer context │ │ Clarify the missing CSV scope │ ├──────────────────────────────────┤ │ Team decides the priority │ │ Next: requirements → roadmap │ └──────────────────────────────────┘
Analyze in the built-in chat or your own assistant
Open the AI Assistant beside your board, or use a connected cloud MCP client. Supply the scope and ask for a proposal before any changes.
Start with the built-in AI chat.
Open the AI Assistant beside your boards. Ask questions about feedback, draft requirements or configure a workflow, then review the result with your team. You can start here without connecting an external MCP client.
See the built-in AI Assistant →Connect through cloud MCP.
Work with Ducalis from a compatible AI client such as Codex or Claude Code. Use it from your desktop, or your phone when the client supports remote MCP. Add the server URL from your API, MCP settings and authorize the connection in your browser through OAuth.
See MCP connection steps →Try this: “Analyze the selected feedback from the last 30 days. Propose groups by customer problem, link every conclusion to source requests, separate repeated mentions from distinct customers and list open questions. Do not merge, create or update records.”
The assistant works with accessible Ducalis context such as Ideas, issues and comments. Bring external reviews into the selected input or supply the excerpts you want analyzed. Check the coverage of a large analysis and split it into scoped batches when needed.
Explore AI Assistant context and access →Group by the problem, then check the meaning
Different wording can describe the same customer outcome. Ask AI to group requests by the problem and intended result, with a short explanation and source links for each proposed theme.
Keep close but distinct requirements separate. “Export invoices as CSV” and “Email invoices every week” concern invoices, but they require different product behavior.
Review duplicate candidates separately from thematic groups. After confirming the same need, use Ducalis’s Merge ideas flow to choose the Idea that should remain. Grouping a theme does not itself merge records or Jira and Linear issues.
┌─ CHECK AGAINST THE SOURCES ──────┐ │ Read each original request │ │ Confirm matching needs │ │ Keep distinct requests apart │ ├──────────────────────────────────┤ │ Approve theme and context │ │ Review duplicates separately │ └──────────────────────────────────┘
Example: four messages, two distinct needs
Illustrative source records. The customer identifiers are supplied, rather than inferred from writing style.
| Source | Supplied feedback | Proposed analysis |
|---|---|---|
| R17 · Acme · widget | “I need to export invoices as CSV for my accountant.” | CSV-export theme; source for Idea CSV-17. |
| R23 · Acme · Slack | “Following up on our CSV-export request so we can share invoices.” | Another mention from the same supplied account; not another customer. |
| R31 · Birch · Discord | “Can I download invoice data as a CSV for our finance team?” | Similar outcome from a second account; confirm required columns. |
| R44 · Delta · supplied review | “Please send invoices by email every week.” | Weekly-email need; keep separate from CSV export. |
Reviewed conclusion
R17, R23 and R31 describe a need to share invoice data outside the product. Keep the evidence under the CSV-export theme and confirm the export scope.
Open questions
Which columns and date filters are required? Can the two accounts use the same export? R44 asks for recurring email delivery and needs its own requirements.
Make recurring problems useful for planning
Keep the scope, source evidence and missing information visible in the analysis.
Keep the original evidence
Give each theme source links and supporting excerpts. Check the assistant’s summary against the actual customer wording before using it.
Separate counts and context
Repeated posts can come from one account. Use known identifiers to compare mentions and customers; leave missing identity or revenue information unknown.
Use themes as decision inputs
A frequent request is evidence to discuss. Combine it with impact, effort, customer context and product goals when setting backlog priorities.
Link the finding to work your team can prioritize
For the reviewed CSV-export theme, check the existing backlog before creating another task. Link Idea CSV-17 to the illustrative Jira BILL-142 export issue and retain the customer evidence. Use the same reviewed handoff for Linear work.
If no matching work exists, draft a problem statement, requirements and open questions for a new issue. Confirm the destination and verify the saved result in Ducalis.
The team evaluates impact and effort, then decides whether the work belongs in the roadmap. The separate weekly-email request stays visible instead of disappearing into the export theme.
Draft requirements from reviewed feedback → · Evaluate backlog priorities →
┌─ EVIDENCE FOR TEAM PRIORITIES ───┐ │ Review impact and effort │ │ Add known customer context │ │ Clarify the missing CSV scope │ ├──────────────────────────────────┤ │ Team decides the priority │ │ Next: requirements → roadmap │ └──────────────────────────────────┘
Put feedback analysis in context
Explore collection, customer context and feature preferences before turning reviewed findings into priorities.
Start with one set of customer feedback.
Read the sources, review the themes and connect the findings to your next backlog decision.
Analyze feedback with AICollect and organize requests first → · Share the work you decide to build →