Validate email hero images with Gmail, Dropbox, OCR.Space and Google Sheets

How it works

Expected Image Download The expected image’s Dropbox URL is passed directly into an HTTP Request node, which downloads the image as binary data.

Actual Image Extraction The HTML node extracts the src of the image from the email (e.g., hero image or other targeted image sections). This src is passed into another HTTP node to download the actual email image in binary format.

OCR Processing Both binary images are sent separately to OCR.Space API via POST requests. OCR returns extracted text for each image.

Comparison A JS node compares the Expected Image Text vs Actual Image Text. If the texts match, the result is marked Match; otherwise Mismatch.

Write Back to Google Sheets The workflow writes ExpectedText, ActualText, and Result back into the Google Sheet, creating a clear audit trail for validation.

Setup steps

1. Google Sheets Create a sheet with the following columns:

SectionId ExpectedImageURL ExpectedText ActualText (output) Result (output)

Add the expected text that should appear inside each image.

2. Email HTML Source Use one of the following:

Gmail node — fetches the email HTML via unread or filtered messages HTTP Request / binary file input — if HTML is being pulled from storage or a URL Ensure the HTML extraction node can find the image src correctly

3. Expected Image (Dropbox URL)

Paste the public Dropbox URL directly into the Google Sheet This URL is fed into an HTTP Download node The resulting binary is used for OCR extraction

4. OCR Setup

Add your OCR.Space API key in the HTTP POST node Set the request body to include:

apikey the binary property of the image desired OCR settings (language, text extraction mode)

5. Merge & Compare

Use a Merge node to align:

ExpectedImageURL + ExpectedText ActualImageSrc OCR outputs

6. In the Code node:

Normalize both texts Compare and assign Match or Mismatch

7. Write Output Use a Google Sheets Append/Update node to write:

ActualText Result

back into the corresponding row, using SectionId or ExpectedImageURL as the matching key.