Weekly Manychat feedback digest in Notion with a Slack ping

By General Input

Every Friday at 4pm, turn the week's Manychat feedback into a clean product digest in Notion and ping the team in Slack with the top themes.

Integrations

  • Manychat
  • Notion
  • Slack Bot

Type

Agentic Task

Categories

  • Product
  • Customer Support

Every Friday at 4pm in my local timezone, build a weekly product feedback digest from Manychat, write it to Notion, and ping our product team in Slack.

Set up: at the start of the run, call Manychat's Get Custom Fields once to resolve which custom-field IDs hold feedback text. The likely names are feedback_text, review_text, and nps_comment, but read the page's real custom fields and match by name. Also call Manychat's Get Tags so the agent can recognize feedback-signal tags such as feedback, nps, review, or bug_report.

Collect feedback for the last 7 days. For each feedback custom field, use Manychat's Find Subscriber by Custom Field to pull every subscriber whose value is non-empty. Also include any subscribers carrying the feedback-signal tags identified above. De-duplicate by subscriber ID. For every subscriber in the combined set, call Manychat's Get Subscriber Info to fetch the full profile, custom field values, and tags.

Filter to feedback created or updated in the last 7 days based on the subscriber's last_interaction or last_growth_tool timestamps and the custom-field last-updated time when available. If a subscriber's feedback predates this window, drop them.

If the resulting set is empty, exit cleanly without writing to Notion or posting to Slack. Skipping quiet weeks is correct behavior.

Otherwise, analyze the feedback. Categorize every piece into one of: bug, feature request, pricing, onboarding friction, praise, or other. Then cluster recurring themes (e.g. 'onboarding video too long', 'wants Stripe integration', 'pricing confusion on the team plan'). Pick the top 3 themes by how many subscribers mentioned them. For each theme, choose 1-2 representative direct quotes with the subscriber's first name and Manychat handle.

Write the digest using Notion's Create a Page into our product feedback database (ask me for the database ID on first run, then remember it). Use a consistent page format so the database reads like a running log:

- Title: 'Manychat feedback digest, week of <Mon DD>' - Top of page: a 2-3 sentence summary of the week (volume, dominant categories, mood). - Section 'Top 3 themes': for each theme, a heading with the theme name, 1-2 quotes (italicized), the subscriber handles, and a one-line recommended follow-up. - Section 'Category breakdown': counts for bug, feature request, pricing, onboarding friction, praise, other. - Section 'All feedback': a bullet list of every quote with category tag and subscriber handle. - Section 'Recommended follow-ups': 3-5 concrete next steps for the product team.

After the Notion page is created, post to our product Slack channel using Slack Bot's Send a Message. Default channel is #product-feedback (ask me on first run, then remember). The Slack message should be short: one line of context, the top 3 themes as bullets with the subscriber counts, and a link to the Notion page. Keep the message under ~6 lines.

Important constraints: keep the Notion page format identical week over week, do not invent feedback or paraphrase quotes (use direct text from the custom field), and skip the run cleanly with no Notion or Slack output if there is no new feedback. Always include the subscriber handle next to every quote so the team can follow up directly.

Related prompts

Explore more prompts
A brand asset library your marketing team actually searchesTurn Mailjet email clicks into ranked HubSpot follow-upsClean out the Looker dashboards and Looks nobody opensLiveKit live operations console for room moderationWake up dormant Keap leads with a researched reasonLiveChat coverage board for planning next week's shiftsPhone routing control panel for LiveKit voice agentsLinkedIn Ads budget pacing dashboard for every client accountGive your team Looker numbers without buying more seatsPause marketing emails to escalated customers, then restore them