Product planning
Sort feature requests by the problem they describe
Group requests by the task the person is trying to finish, not by the feature they named. Different wording often describes one problem, and one wording can hide two. Record what you know and what you still need to ask.
A text to start with
Here are [number] feature requests from [source, such as support tickets or interviews]: [paste requests with dates and senders]. Group them by underlying user task. For each group, list the original requests, the task as I would state it, any evidence beyond the requests, the users affected if known, and the open question that would clarify it. Do not rank or estimate effort.
Example request. Change the details to fit your situation.Request triage sheet
Request (original wording): [ ] Sent by (role or account and name if needed): [ ] Date and channel: [ ] Underlying task: [ ] Evidence beyond the request: [ ] Users affected (count and how counted, or unknown): [ ] Duplicate of or related to: [ ] Open question: [ ] Who can answer it: [ ] Group name: [ ] Requests in this group: [ ] Distinct people or accounts: [ ] Status (needs context, ready to compare, set aside) and reason: [ ]
Volume versus need
Check whether a group looks large only because one person or account wrote several times, or because one channel gets more attention than others. Count distinct people before counting messages. Confirm the stated task comes from the requests and was not invented to fit a feature you already have in mind. Look for requests merged into one group that describe different tasks, and for duplicates left apart because the wording differs. Where the draft says how many users are affected, confirm the number comes from your data; if it does not, mark it unknown.
Ask another agent to check the result
Check whether any group is described as important mainly because its requests were loud, frequent or from one source. Compare distinct requesters with message counts. Also find groups that combine different tasks and duplicates split across groups. Quote the request text for each finding.
Fill the gaps before planning
Take the groups with the most open questions and decide how to answer each: a follow-up message to the requester, a look at usage data, or a short conversation with support. Add what you learn to the sheet and keep the original requests attached to each group so the source stays visible. Only then compare groups against each other, and record the reasoning for your ordering.