DraftThis page is a draft prepared for review and is not yet legal advice. Items in dashed boxes are placeholders to complete before publishing.
The problem we started with
Every product in a catalogue is described several times. It is in the feed sent to Google, in the structured data search engines read, and on the page shoppers see. Each of those can be checked on its own and pass. They can still disagree with each other, and when they do, it is usually the product, its visibility or its eligibility that pays for it.
Feed validators check the feed. Schema validators check the markup. Crawlers check the page. We built SKUDrift to check whether they agree.
What we believe
Evidence, not a score
Every finding shows what each source actually said and when, so you can see the problem rather than trust a number.
Coverage, stated
We say how much of your catalogue we could compare. What we could not check never quietly counts as a pass.
False positives are defects
A finding that turns out to be wrong wastes your time and your trust. We treat it as a bug, not noise.
Nothing invented
We only describe a Google consequence where Google documents it. No scare stories, no guesswork dressed as a rule.
How we build it
SKUDrift is built by a small team that has spent years inside product feeds, structured data and Merchant Center accounts. We are building it with a small group of ecommerce teams, agencies and consultants who work with product feeds every day, and who tell us where their data really goes wrong.
Where we are
SKUDrift is in development. It starts with Google Shopping, with ChatGPT support coming soon, and it is designed for a future where AI agents compare and recommend products using the same data.
Help shape SKUDrift.
All you need is the feed you already send to Google.