Designing the Product Analysis Workflow

Your workflow takes Amazon ASINs from a spreadsheet, fetches product data via API, processes it with AI, and outputs analysis reports. Part 1 focuses on setup, data-validation paths, and HTTP request configuration.

A workflow that:

• Reads ASINs (Amazon product identifiers) from a spreadsheet;
• Validates whether each ASIN exists;
• Fetches product data from the Rainforest API;
• Checks whether the API returned usable data;
• Sends the product data to ChatGPT for analysis;
• Generates a summary and improvement suggestions;
• Creates a Google Doc containing the report;
• Links the document back to the original spreadsheet row.

# Spreadsheet Row (ASIN)
#     ↓
# ┌─────────────────────────────────────────┐
# │              PATH: ASIN Check           │
# ├──────────────────┬──────────────────────┤
# │   No ASIN        │    ASIN Found        │
# │      ↓           │         ↓            │
# │ Update Sheet:    │   HTTP Request       │
# │ "ASIN not found" │   (Rainforest API)   │
# │                  │         ↓            │
# │                  │ ┌─────────────────┐  │
# │                  │ │  PATH: Data     │  │
# │                  │ │     Check       │  │
# │                  │ ├────────┬────────┤  │
# │                  │ │Poor    │Good    │  │
# │                  │ │Data    │Data    │  │
# │                  │ │   ↓    │   ↓    │  │
# │                  │ │Update  │ChatGPT │  │
# │                  │ │Sheet:  │   ↓    │  │
# │                  │ │"Poor   │Google  │  │
# │                  │ │ data"  │Doc     │  │
# │                  │ │        │   ↓    │  │
# │                  │ │        │Update  │  │
# │                  │ │        │Sheet   │  │
# └──────────────────┴────────┴────────┴───┘

The workflow starts with a Google Sheets trigger that runs when an ASIN is added or updated. Using ASIN as the trigger column gives precise control over when processing begins and ensures each row is handled intentionally. A quick trigger test confirms the row number, ASIN, and related fields are passed correctly.

Before any API calls, the workflow uses Paths for validation. If the ASIN is missing, the row is immediately updated with a failure status and processing stops. If an ASIN exists, the workflow continues and makes an HTTP GET request to the Rainforest API, dynamically mapping the ASIN from the spreadsheet. This step fetches full product data, which becomes available for downstream logic.

Because API responses are not guaranteed to be usable, a second data-quality Path checks whether a product title exists. Missing or incomplete data is flagged back in the spreadsheet, creating a clear audit trail. When valid data is present, the workflow sends the full API response to ChatGPT for analysis, setting up AI processing that will generate summaries and recommendations in the next part.