How it works.
One product page per audit. Most take 10 to 30 seconds.
Paste a product URL
Optionally add the product's row from your feed, so the feed is checked against the page as well.
We read the page
Code checks price, stock, identifiers, Product markup and crawl rules. AI judges how well each attribute is explained, quoting the page.
Fix what matters first
You get a score, the top fixes, and a row for every attribute. Copy it as Markdown, download a CSV, or share a link.
Your feed is a checklist.
If an attribute matters enough to send Google as a structured field, it matters enough to explain on the page. These eight rows come first in every report.
| Feed field | Where it lives on the page | What “detailed” looks like |
|---|---|---|
title | H1 | Brand, product type and the attributes people search for, matching the feed title closely |
description | Main product copy | What it is, who it's for, what problem it solves, written for a shopper, not a spreadsheet |
product_highlight | Bullets near the buy button | The few things that actually decide the purchase |
product_detail | Specs table | Every technical attribute, with units |
color, size, material | Variant selector and specs | Size guide, fit notes, material and care details |
price, sale_price, availability | Buy box | Identical to the feed at all times |
shipping, return_policy_label | Buy box and a shipping/returns section | Delivery times and costs by region, plus the return window, in plain words |
image_link, additional_image_link | Gallery | Multiple angles, scale and in-use shots, with descriptive alt text |
Method by Gagan Ghotra. The report also covers brand, condition, dimensions, who it's for, pattern and compatibility where they apply.
What it catches.
The issues behind Merchant Center disapprovals and lost rich results, plus the content gaps that lose the sale.
Mismatches
Price, sale price, availability, condition, GTIN and brand that differ between the page, its markup and your feed.
Markup and crawl problems
Missing Product fields, invalid GTINs, variant prices missing from markup, noindex, robots.txt blocks and JavaScript-only content.
Thin product content
Attributes your feed sends but the page never explains, like delivery times, the return window, specs with units and size guides.
Questions.
What does the Feed-to-Page Checker do?
It reads one product page and checks it two ways. First, that the page, its Product structured data and your Merchant Center feed agree on price, availability, condition and identifiers, the mismatches that get products disapproved. Second, that the page actually explains what your feed sends Google: title, description, specs, colour, size, material, shipping, returns and more, with a quote from the page for every verdict.
Do I need my feed?
No. Without it, the tool checks the page against its own markup and judges every attribute from the page. If you paste the product's row from your feed (TSV, CSV or one XML item), it also checks that the feed matches the page.
How much does it cost?
Your first audit is free with an account. Audit packs, paid once with no subscription, are coming soon. If an audit fails because we couldn't read the page, the credit is refunded.
How is AI used?
Prices, stock, identifiers, markup and crawl checks are exact code checks. Only the question of whether the page explains an attribute well is judged by an AI model (Claude), and it must quote the page. We check every quote against the page text and flag any we can't find.
Does it work on any store?
Yes: Shopify, WooCommerce and custom stores. It reads the HTML Google downloads first and doesn't run JavaScript yet, so it tells you when key details only appear after scripts run.
Whose method is this?
It's based on Gagan Ghotra's approach to product page SEO: your Merchant Center feed is already a checklist for the product page. If something matters enough to send Google as a structured field, it matters enough to explain on the page.
More tools.
Free and private: they run in your browser on your Search Console export.
Hear what your data isn't saying.
One email a week with the findings, experiments and tools we ship.