Amazon Review Analysis using Gemini AI and Google Sheets - n8n Workflow

Markdown-formatted list:
Product Managers seeking detailed, actionable customer feedback (Voice of Customer).
E-commerce sellers (Amazon FBA/D2C) who need to rapidly monitor product reputation.
Data analysts looking for automated methods to structure qualitative review data.
Anyone needing robust n8n templates for web scraping and AI processing.

Manually sifting through thousands of Amazon reviews is slow and inefficient. This specialized n8n workflow template solves this by providing a hands-free automation pipeline.

It leverages Apify to scrape relevant reviews (focused on 1 and 2-star feedback), then sends each review through the Gemini large language model for advanced categorization, sentiment scoring, and concrete improvement suggestion generation. The structured output is then instantly written to a Google Sheet, providing a clean database of actionable insights for your product team.

This robust n8n workflow ensures you never miss critical feedback, transforming raw text into valuable business intelligence. Setting up this n8n automation is key for efficient market research.

This powerful n8n automation follows a five-stage process, starting with a simple n8n trigger:

1. Trigger and Setup: The n8n workflow begins with a Manual Trigger, followed by a 'Set' n8n node to define key variables like the maximum number of reviews to scrape (maxReviews) and the Apify Actor ID.


2. Data Extraction (Apify): The Apify n8n node executes the Amazon Reviews Scraper actor, fetching low-rated reviews from the target Amazon product URL.


3. Batch Processing and AI: The 'Split In Batches' n8n node processes each extracted review individually. Each review item is then routed to the Google Gemini n8n node. The AI uses a detailed prompt to assign a numerical sentiment score (1-5), categorize the issue, summarize the complaint, and suggest a specific improvement.


4. Data Structuring: A Code n8n node cleans and parses the text output from the AI, converting it into a clean JSON object ready for database insertion. This is crucial for reliable data handling in this n8n workflow.


5. Storage and Notification: The structured data is appended as a new row in a configured Google Sheet. Once all items are processed, the batch loop completes, and a final Slack n8n node sends a notification confirming the analysis is finished, detailing the number of reviews processed by the n8n workflow.

To deploy this n8n workflow template, follow these steps:

1. Import: Copy the provided JSON into your n8n instance (Cloud or Self-hosted) via the 'New' > 'Import from JSON' option.


2. Credentials: Set up credentials for:

1. Configuration: In the 'Run an Actor' Apify n8n node, update the startUrls parameter to the specific Amazon product URL you wish to analyze. Verify the columns in your target Google Sheet match the expected JSON output (sentiment_score, category, summary, improvement).


2. Activation: Once all credentials and configurations are set, activate the n8n workflow and execute the Manual Trigger.

This n8n workflow leverages several specialized n8n node types:

Manual Trigger: The starting n8n trigger for manual execution. It allows immediate initiation of the n8n workflow.
Workflow Configuration (Set n8n node): Stores critical settings like maxReviews (defaulted to 10) and apifyActorId.
Run an Actor and get dataset (Apify n8n node): Responsible for the web scraping component, extracting Amazon reviews based on specified rating filters (1 and 2 stars).
Loop Over Reviews (Split In Batches n8n node): A core logic n8n node that handles iterative processing, ensuring each review is analyzed individually by the AI.
Message a model (Google Gemini n8n node): The brain of the operation. This n8n node sends review text to Gemini 2.0 Flash Lite, using a sophisticated prompt to force structured JSON analysis of sentiment and actionable improvements.
Code (Parse JSON)1 (n8n Code node): A custom scripting n8n node that cleans the AI's textual response, removing markdown artifacts (like ```json), and converts the resulting string into a usable JSON object.
Create spreadsheet (Google Sheets n8n node): Appends the highly structured, AI-generated data into a Google Sheet row-by-row.
Slack - Send Completion Notification (Slack n8n node): Provides final confirmation that the n8n workflow has executed successfully, including metrics on the batch size processed.