n8n Workflow: Prioritize Feature Requests with Claude + Slack

Turn customer feedback into an explainable, prioritized roadmap—automatically

This n8n workflow batches incoming customer feedback via webhook, uses Anthropic Claude to extract and theme real feature requests, then calculates a transparent 0–100 priority score and delivers a ranked report to Google Sheets, Slack, and SendGrid.

What this workflow does

• Receives feedback: Accepts a webhook POST containing an array of raw feedback items (or falls back to a built-in sample payload if none is provided).
• Identifies feature requests: Uses Anthropic Claude to classify each item as a feature request and extract structured fields such as title, category, sentiment, and urgency.
• Clusters into themes: Filters out non-feature items, then aggregates extracted requests into themes using Claude—complete with mention counts, customer reach, and averaged urgency/sentiment scores.
• Computes an explainable priority score: For each theme, generates a deterministic 0–100 priority score, assigns a tier (Critical/High/Medium/Low), and suggests timing based on mentions, reach, tiers, urgency, and sentiment.
• Generates a leadership-ready report: Claude produces an executive summary and a ranked list of themes grounded in the computed scores.
• Delivers results: Appends a summary row and the full ranked theme JSON to Google Sheets, sends a Slack alert when any theme is Critical, emails a digest via SendGrid, and returns the report as the webhook response.

Use cases

• Capture SaaS feature requests from support, surveys, or in-app feedback and convert them into a prioritized product review.
• Rapidly spot “Critical” themes and notify your team in Slack with an evidence-based ranking.
• Maintain a historical tracker of feedback themes in Google Sheets with structured JSON for auditability.

Technical details

• Workflow control: Uses nodes such as if, set, code, wait, webhook, split out.
• AI integration: Configure an Anthropic API credential and use it for three Anthropic Claude chat model steps (classification/extraction, theme clustering, and report generation).
• Outputs: Google Sheets logging, Slack critical alerts, and SendGrid email digest.