Automate Amazon Review Analysis with n8n & Bright Data

1. Opening Problem Statement

Meet Sarah, an e-commerce brand manager who spends several hours every week manually digging through thousands of Amazon product reviews to understand what customers like and dislike about her competitors’ products. This painstaking process of collecting and analyzing reviews is not only time-consuming but also prone to errors and oversight, leading to delays in campaign adjustments and lost opportunities to outshine competitors with targeted marketing.

Imagine Sarah’s frustration: she wastes 5-8 hours weekly scraping review data, sorting through it in spreadsheets, and trying to extract meaningful insights. Mistakes creep in from manual data entry, and vital negative feedback or trends get overlooked. These inefficiencies directly cut into her marketing budget and slow her ability to innovate.

This is exactly where our n8n workflow comes to the rescue, automating the entire process of Amazon review scraping, summarizing feedback using AI, and generating creative content for media buyers—all seamlessly integrated.

2. What This Automation Does

This n8n workflow streamlines competitor review analysis on Amazon by:

• Triggering on-demand via a simple form input where you paste an Amazon product URL.
• Using Bright Data’s API to scrape all reviews for the specified product automatically.
• Polling until the scraped data is ready, ensuring real-time processing without manual refresh.
• Importing the raw review data into a Google Sheets template for easy storage and visualization.
• Aggregating and summarizing review texts using OpenAI’s GPT-4 model to highlight competitor weaknesses.
• Generating a custom, weird-and-fun AI image ad creative based on review insights that media buyers can immediately use in campaigns.
• Sending the AI-generated summary and image as an email to stakeholders, facilitating quick decision-making.

Sarah now reclaims 6+ hours per week, eliminates manual errors, and gains sharper competitive intelligence that perfectly targets marketing pain points.

3. Prerequisites ⚙️

• n8n Automation Platform (cloud or self-hosted) 🔌
• Bright Data API Key for Amazon review data scraping 🔑
• Google Sheets Account with OAuth2 credentials for data storage 📊
• OpenAI API Key to leverage GPT-4 for review summarization and image creation 🔐
• Gmail Account with OAuth2 credentials for sending summary emails 📧

If you prefer self-hosting n8n for full control, consider providers like Hostinger for easy setup.

4. Step-by-Step Guide to Build This Workflow

Step 1: Set Up the Form Trigger for Amazon Product URLs

Navigate to Triggers and add the Form Trigger node.
Configure the form field labeled “Amazon Product URL” as required and use the example placeholder shown in the workflow. This node will start the flow when you paste in a product URL.
You should see a webhook URL generated—this is where the form submissions are received.
Expected outcome: The workflow listens for submitted Amazon URLs to process.

Common mistake: forgetting to make the URL field required or misconfiguring the webhook.

Step 2: Use HTTP Request to Trigger Bright Data API

Add the HTTP Request node named “Post API call to Bright Data.”
Set Method to POST.
Enter the URL: https://api.brightdata.com/datasets/v3/trigger.
Under Query Parameters, add dataset_id with the ID gd_le8e811kzy4ggddlq (this dataset collects Amazon reviews).
In the JSON body, map the product URL from the form trigger: [{"url":"{{ $json['Amazon Product URL'] }}"}].
Add the header Authorization with your Bright Data Bearer token.
This sends the scraping request to Bright Data.
Expected result: A snapshot ID for data retrieval is received.

Common mistake: missing API key or incorrect JSON formatting.

Step 3: Implement a Wait Node for Polling

Drag in a Wait node set to 1 minute.
Connect the HTTP Request node output here.
This introduces a delay giving time for Bright Data to start scraping.
Expected outcome: The workflow pauses and retries status checks regularly.

Step 4: Check Snapshot Progress Using HTTP Request and If Node

Add another HTTP Request node “Snapshot Progress” to poll the Bright Data status endpoint, using the snapshot ID received.
URL: https://api.brightdata.com/datasets/v3/progress/{{ snapshot_id }}.
Configure Authorization header similarly.
Then add an If node to check if the status equals “running” (meaning snapshot is still processing).
If true, loop back to the Wait node for re-polling; otherwise continue.
Expected outcome: The workflow only proceeds when data is fully ready.

Common mistake: not looping correctly causing the workflow to fail prematurely.

Step 5: Retrieve the Completed Snapshot Data

Add HTTP Request node “Getting data from Bright Data” to fetch the scraped review JSON.
URL: https://api.brightdata.com/datasets/v3/snapshot/{{ snapshot_id }}, add format=json query.
Send authorization header.
Expected output: Full review data in JSON format.

Step 6: Append Reviews to Google Sheets

Use the Google Sheets node configured with your spreadsheet ID (provided template link is included in Sticky Notes).
Operation: Append.
Map the relevant review fields such as review_text, author_name, rating, and others automatically.
Expected result: Structured review data appears in your Google Sheets for analysis and backup.

Common mistake: mismatched field mapping or wrong spreadsheet ID.

Step 7: Aggregate All Review Texts

Add the Aggregate node to combine all review texts into a single string field.
This prepares the data for AI summarization.
Expected outcome: One aggregated text of all reviews.

Step 8: Summarize Reviews Using OpenAI GPT-4

Insert the Basic LLM Chain node (LangChain) with a prompt asking:
“Read the following reviews, these are reviews of our competitors: [aggregated reviews]. After reading them, summarize their weakest points. Don’t mention the competitor name.”
Model used: GPT-4o-mini.
Expected output: Clear, insightful summary of competitor product weaknesses.

Step 9: Generate AI Image Creative with OpenAI

Add the OpenAI - Generating Image node.
Use a JSON prompt specifying ad dimensions (1080×1080), target audience (B2C), “weird-and-fun” style, and focus on the biggest competitor pain point.
Expected output: Unique marketing image to use in digital ads.

Step 10: Email Summary and Creative to Media Buyers Using Gmail

Add the Gmail node.
Configure recipient email (example in workflow: [email protected]).
Include the summarized review text and attach the AI-generated image.
Expected outcome: Stakeholders receive actionable insights and creative content automatically for speedy campaign deployment.

5. Customizations ✏️

• Change Review Data Storage: In the Google Sheets node, you can modify the spreadsheet ID or sheet name to use your own document for storing reviews.
• Adjust AI Summary Prompt: Modify the text field in the Basic LLM Chain node to add your company details or refine the summary focus, such as highlighting positive features instead.
• Polling Interval Tuning: Edit the Wait node duration to poll Bright Data faster or slower based on dataset size and API limits.
• Email Recipient Customization: Change the recipient email in the Gmail node to notify different team members or integrating into Slack instead.
• Image Style Tweaks: Alter the “weird_objects” array or color scheme in the OpenAI Image node prompt to align the ad creative with your brand personality.

6. Troubleshooting 🔧

Problem: “HTTP Error 401 Unauthorized” when calling Bright Data API.
Cause: Incorrect or missing API key.
Solution: Verify your Bright Data API token in the HTTP Request node headers and ensure it is current.

Problem: “Snapshot status stuck at running forever.”
Cause: Bright Data API delay or erroneous snapshot ID.
Solution: Check the snapshot ID passed, increase wait polling time. Consider retry logic or manual snapshot cancellation if needed.

Problem: Google Sheets node fails to append data.
Cause: Incorrect document ID or OAuth2 token expired.
Solution: Confirm Google Sheets credentials, re-authenticate and ensure the sheet ID and permission scopes are correct.

7. Pre-Production Checklist ✅

• Confirm Bright Data API access and dataset ID validity.
• Test form trigger webhook using sample Amazon URLs.
• Verify Google Sheets integration with test data append.
• Ensure OpenAI API keys and quota are active for GPT-4 and image generation.
• Test email sending with valid Gmail OAuth2 credentials.
• Run end-to-end tests to confirm snapshot polling logic and data flow.
• Backup your Google Sheets data in case of unexpected overwrite or errors.

8. Deployment Guide

Activate the workflow in your n8n environment by toggling it live.
Use the generated form URL to start new scraping jobs by submitting competitor Amazon product links.
Monitor execution logs in n8n for any failures or long polling times.
Leverage the Slack or email notifications to alert you on errors (optional add-on).
Keep your API keys updated for uninterrupted scraping and summarization.

9. FAQs

Q: Can I use other scraping services besides Bright Data?
A: Yes, but the workflow is specifically tailored to Bright Data’s API and dataset ID. You’d need to adjust the HTTP requests accordingly.

Q: Does this consume OpenAI API credits?
A: Yes, GPT-4 summarization and image generation use your OpenAI quota, so monitor usage to avoid unexpected costs.

Q: Is my data secure during transfer?
A: The workflow uses HTTPS secured APIs and OAuth2 authentications to protect your data. Ensure your API keys are managed safely.

Q: Can it handle multiple Amazon URLs at once?
A: The form trigger handles one URL per submission, but you can run multiple jobs in parallel by submitting multiple forms.

10. Conclusion

By building this customized n8n workflow, you’ve automated the end-to-end process of scraping, aggregating, analyzing, and creatively visualizing Amazon competitor reviews. Sarah’s weekly 6-hour manual effort is now condensed into minutes, with added accuracy and actionable insights delivered directly to her team.

This automation empowers your marketing strategy with data-driven creative content, saving time and increasing conversions. Next steps could include integrating sentiment analysis, automating ad placement on Facebook, or extending support to other e-commerce platforms for a holistic view.

Start implementing this workflow today and transform how you leverage customer feedback for competitive advantage.