n8n templatesn8n templates

Trustpilot Customer Review Insights Generator

This workflow automates the scraping and analysis of customer reviews for specified companies on Trustpilot. It utilizes a vector database for storage and similarity search, combined with the K-means clustering algorithm to group similar feedback. Advanced natural language processing techniques are employed to generate detailed customer insights and sentiment analysis reports, which are then exported to Google Sheets for easy team analysis and sharing. This process efficiently identifies customer opinions, aiding in market research, customer service, and product improvement, ultimately enhancing customer satisfaction.

Workflow Diagram
Trustpilot Customer Review Insights Generator Workflow diagram

Workflow Name

Trustpilot Customer Review Insights Generator

Key Features and Highlights

This workflow automatically scrapes Trustpilot review data for a specified company, leverages the advanced vector database Qdrant for storage and similarity search, applies the K-means clustering algorithm to group review content, and utilizes the OpenAI GPT-4 model to generate detailed customer insights and sentiment analysis reports. Finally, the results are exported to Google Sheets for easy further analysis and sharing.

Core Problems Addressed

  • Automating the extraction and structured processing of large volumes of customer reviews
  • Quickly identifying and grouping similar customer feedback themes
  • Deeply mining key insights and sentiment tendencies from customer reviews through natural language processing
  • Efficiently generating actionable customer insight reports to enhance customer satisfaction and product improvement efficiency

Application Scenarios

  • Market research teams analyzing customer feedback for competitors or their own brand
  • Customer service departments monitoring customer satisfaction and pain points
  • Product managers obtaining user experience feedback to guide product optimization
  • Brand marketers extracting reputation information to formulate marketing strategies

Main Workflow Steps

  1. Set Target Company: Enter the company identifier from the corresponding Trustpilot URL.
  2. Clear Old Data: Use the Qdrant API to clear historical vector data for the company, ensuring data freshness.
  3. Scrape Review Data: Paginate through recent Trustpilot review pages via HTTP requests.
  4. Parse Review Content: Extract review author, rating, title, content, date, country, and other details using HTML node parsing.
  5. Data Integration and Storage: Convert review data into vectors and store them in the Qdrant vector database.
  6. Trigger Sub-Workflow: Select the desired time range to retrieve relevant review vectors.
  7. Clustering Analysis: Execute K-means clustering on vector data using a Python code node to group similar reviews.
  8. Retrieve Cluster Details: Extract specific review data within each cluster.
  9. Generate Customer Insights: Call the OpenAI GPT-4 model to summarize each cluster’s reviews, perform sentiment analysis, and provide improvement suggestions.
  10. Export Results: Append insights and raw data to Google Sheets for team review and subsequent processing.

Involved Systems and Services

  • Trustpilot: Data source for customer reviews.
  • Qdrant Vector Database: Stores and retrieves review vectors, enabling efficient similarity search and clustering.
  • OpenAI GPT-4: Performs natural language processing to generate insight reports and sentiment analysis.
  • Google Sheets: Stores and shares analysis results.
  • n8n: Workflow automation platform integrating all services to achieve end-to-end automation.

Target Users and Value

  • Data Analysts and Market Researchers: Quickly obtain structured customer feedback insights to improve analysis efficiency.
  • Customer Service and Product Teams: Discover common issues and product strengths/weaknesses through review clustering.
  • Business Managers and Decision Makers: Gain real-time customer voice insights to formulate precise improvement actions.
  • Automation Enthusiasts and Developers: Learn how to combine vector databases and large language models for intelligent data analysis.

This workflow is especially suited for handling massive volumes of customer reviews, helping enterprises extract valuable insights from complex textual data and achieve data-driven customer experience optimization.