Customer Feedback Analyzer

Weekly Customer Intelligence From Every Feedback Channel at Once

Customer feedback is scattered across G2, the App Store, Intercom, Typeform surveys, and support tickets — and synthesizing it manually into something actionable takes hours. Most teams don't do it consistently, so they miss the patterns: the same complaint appearing in three channels, the feature request that came up seven times, the sentiment shift that started three weeks ago. The Customer Feedback Analyzer aggregates all of it, identifies themes, tracks sentiment trends, and delivers a weekly executive summary automatically.

What It Does

This workflow runs on two schedules: every 12 hours for critical feedback collection and alerting, and every Monday at 8 AM for the full weekly digest.

Data Collection (every 12 hours):

• G2 Reviews: Fetches recent reviews from your G2 product page, parsing star ratings, review text, use case, and company size
• Intercom Conversations: Pulls recent support conversations and analyzes complaint patterns and sentiment
• App Store Reviews (configurable): Fetches and parses ratings and review text from your Apple App Store listing
• Typeform Surveys (configurable): Processes survey responses from any connected form

AI Analysis: Claude processes all collected feedback to:

• Identify recurring themes and cluster them by feature area (onboarding, performance, mobile, integrations, pricing, etc.)
• Detect sentiment trends — is sentiment improving or declining, and in which areas?
• Surface the most specific, actionable complaints (not just "the product is slow" but "the export function takes >30 seconds on datasets over 5,000 rows")
• Flag any feedback that signals high churn risk or enterprise customer dissatisfaction

Critical Alerts (when triggered): If sentiment drops significantly week-over-week (configurable threshold, default -10 points), an immediate alert is sent to your configured channels before the weekly digest.

Weekly Executive Summary (Monday 8 AM):

• Top 3 themes by frequency with representative quotes
• Sentiment score and week-over-week trend by feature area
• G2 rating trend
• Specific feature requests that came up 3+ times
• Recommended product priorities based on feedback volume and severity
• Delivered to Telegram and/or Slack (#product channel)

Who This Is For

Product managers, founders, and customer success teams at B2B SaaS companies with enough feedback volume to benefit from automated synthesis. Most effective with 20+ reviews/tickets per week across channels. Setup takes 45 minutes because connecting multiple data sources requires configuration — but the ongoing value is significant.

What's Included

• customer-feedback-analyzer.yml — complete workflow with dual cron triggers, multi-source collection (G2, App Store, Intercom, Typeform), AI analysis, trend tracking, and multi-channel delivery
• Configurable feature area taxonomy for theme classification
• Sentiment alert threshold configuration
• Multi-channel delivery (Telegram + Slack)

Setup Steps

1. Install the workflow — copy customer-feedback-analyzer.yml into your OpenClaw workspace
2. Configure data sources — enable/disable G2, App Store, Intercom, and Typeform; add your product slug and API credentials
3. Define your feature areas — customize the feature_areas list to match your product's actual modules/features
4. Set delivery channels — configure your Telegram handle and/or Slack channel for reports
5. Set alert thresholds — adjust sentiment_alert_threshold based on your baseline
6. Register both cron jobs — the workflow has two schedules; register both with OpenClaw
7. Run manually first to validate data source connections before relying on the schedule

Why This Beats Manual Feedback Review

Reading G2 reviews is something most teams do quarterly, if at all. Intercom tickets get reviewed by support, but nobody's connecting them to G2 patterns. Survey results sit in Typeform. The feedback is there — it's the synthesis that's missing.

This workflow replaces the "we should review our feedback more consistently" goal with an automated system that does it every week. Product teams that use it consistently report catching feature issues 2–3 weeks earlier than they would have otherwise.

Result: product decisions informed by actual customer patterns, not the loudest voice in the room.