Support/First-party starter playbook

Weekly product feedback digest

Turn a week of Zendesk tickets into a ranked, quote-backed feedback digest in Notion.

// what it does
  1. 01Gather the week. Search Zendesk for tickets in scope from the last 7 days and read each ticket's subject, description, comments, tags, and the requester's organization. Bound the run to 200 tickets and say so if there were more.
  2. 02Pull out the feedback. From each ticket, extract feature requests, bug reports, confusing workflows, and praise. Ignore pure how-to questions that the help center already answers, and billing or account admin requests.
  3. 03Group into themes. Merge feedback that describes the same need, even when customers phrase it differently, and assign each theme to a configured product area. Keep a theme separate when you are not sure it is the same need.
  4. 04Rank the themes. Order by number of distinct customers, then by ticket count, then by severity (blocked workflow before inconvenience). Note any theme that is new this week or growing compared with last week's digest.
  5. 05Write the digest. Create a Notion page in the configured location with, for each theme: a one-line summary, customer and ticket counts, two or three short direct quotes, and links to the source tickets.
  6. 06Notify the team. Post the top five themes to the Slack channel with a link to the full Notion digest.
// guardrails in the draft

Every theme must link to real tickets, and every quote must be copied from a ticket, trimmed but not reworded.

Remove names, email addresses, phone numbers, and account identifiers from quotes before they leave Zendesk.

Do not reply to, tag, or change any Zendesk ticket. This agent only reads support data.

Do not turn one loud customer into a trend: show customer counts next to ticket counts.

If last week's digest cannot be found, skip the new-or-growing comparison and say so.

Make weekly product feedback digest yours.

Request access and we'll start with a private draft. You keep control of the tools, permissions, tests, and publication decision.