Scrape Trustpilot Reviews with DeepSeek, Analyze Sentiment with OpenAI

For Trustpilot, this automated workflow scrapes reviews, extracts key information, and analyzes sentiment using OpenAI, streamlining the process of gathering customer feedback. It efficiently organizes data into Google Sheets, enabling quick insights into customer opinions and enhancing decision-making.

7/4/2025
20 nodes
Complex
2vg6rbmudj2vzbrjpatcf5qzdjsc2vkymanualcomplexsplitoutlangchainsticky notegooglesheetsadvancedlogicconditionalapiintegration
Categories:
Data Processing & AnalysisManual TriggeredComplex Workflow
Integrations:
SplitOutLangChainSticky NoteGoogleSheets

Target Audience

This workflow is designed for:
- Business Owners looking to analyze customer feedback on Trustpilot to improve services and products.
- Data Analysts who need to extract and analyze sentiment from customer reviews efficiently.
- Marketing Professionals aiming to understand customer opinions and enhance brand reputation through insights from reviews.
- Developers interested in automating the scraping and analysis of online reviews for various applications.

Problem Solved

This workflow addresses the challenge of managing and analyzing customer feedback from Trustpilot. It automates the process of:
- Scraping reviews from Trustpilot, saving time and effort compared to manual collection.
- Extracting key information such as author names, ratings, and review texts, ensuring that important data is captured accurately.
- Analyzing sentiment to gauge customer satisfaction and identify areas for improvement, enabling businesses to respond effectively to customer needs.

Workflow Steps

  • Manual Trigger: The workflow starts when a user manually initiates the process.
    2. Set Parameters: Configuration of parameters like company_id and max_page to define the target company and the number of pages to scrape.
    3. Get Reviews: An HTTP request is sent to Trustpilot to fetch reviews based on the specified company ID.
    4. Extract Review Links: The workflow extracts review links from the HTML content received.
    5. Limit Results: The number of reviews to process is limited to a maximum of 3.
    6. Check Existing Data: It checks if the review has already been saved in Google Sheets to avoid duplicates.
    7. Get Single Review: For each review link, a request is made to fetch the full review content.
    8. Extract Review Information: The workflow extracts relevant data such as author, rating, date, title, text, and user location from the review.
    9. Sentiment Analysis: The extracted review text is analyzed for sentiment (Positive, Neutral, Negative) using AI models.
    10. Update Google Sheets: Finally, the workflow updates a Google Sheets document with all the extracted data and sentiment analysis results, ensuring that the information is organized and easily accessible.

    • Customization Guide

    Users can customize this workflow by:
    - Changing the Company ID: Update the company_id parameter in the Set Parameters step to target a different company on Trustpilot.
    - Adjusting the Maximum Pages: Modify the max_page parameter to scrape more or fewer pages based on needs.
    - Modifying Extraction Criteria: Alter the extraction templates in the Information Extractor node to capture different attributes from the reviews.
    - Customizing Sentiment Categories: Adjust the sentiment analysis categories to fit specific business requirements or focus areas.
    - Integrating Additional Nodes: Users can add more nodes for further processing, such as sending notifications or integrating with other data sources.