Turning Public Customer Feedback into Product and Support Actions

1. Overview

This project is an AI-powered Product + Support workflow that converts raw public customer feedback into two operational outputs:

1. A weekly internal product pulse for the product team
2. A reusable customer-support clarification for a recurring payment-related confusion

The workflow analyzes recent ChatGPT Android Google Play reviews, identifies recurring themes and evidence, generates structured outputs, and requires human approval before it writes to internal tools.

Workflow built with: n8n, Gemini 3.5 Flash-Lite, Google Drive, Google Docs, Gmail, and official OpenAI Help Center sources.

2. Problem and Opportunity

Product and Support teams receive large volumes of customer feedback, but manually reading, organizing, and converting reviews into action is slow and inconsistent.

For a product like ChatGPT Android, feedback may include positive sentiment, feature requests, performance limitations, and confusion around paid features. If these signals are not structured, product teams can miss important patterns and support teams may respond inconsistently.

This workflow addresses that gap by turning reviews into a traceable weekly update and a standardized support explanation grounded in official documentation.

3. Goal

Build an approval-gated AI workflow that:

• Analyzes recent public ChatGPT Android reviews
• Clusters feedback into a maximum of five themes
• Identifies the top three product themes
• Extracts real supporting user quotes
• Detects recurring confusion related to a fee, charge, subscription, or payment