How we check

The method

1. Read Shopify’s version. For each product we pull the summary, “top features,” “tech specs” and selling points Shopify’s catalog gives AI agents like ChatGPT, Copilot and Muse.
2. Read yours. Your description, tags, product details fields and product page, word for word.
3. Split into claims. Each line of Shopify’s version is one claim: the summary, and each of its top features, tech specs and selling points.
4. Compare each claim. AI reviewers check every claim in Shopify’s version against everything you wrote, then sort it into a finding type, a claim category and a severity.
5. Quote, don’t paraphrase. Every finding shows both texts exactly. “…” marks a cut. The claim in question is highlighted and bracketed, never reworded.

The free scan checks 5 of your products, the first 5 Shopify’s catalog returns for your store, and counts how many of your products are in Shopify’s AI catalog. It doesn’t check the rest, and it never shows counts for products it didn’t check. Installing checks every product your plan covers.

Finding types

Always in this order, because this is the order worth fixing in.

Contradiction

Shopify’s version says the opposite of yours. “Fragrance-free” when you list lavender oil.

Invisible to AI

The product isn’t returned by Shopify’s AI catalog at all, so there’s nothing to compare.

You never said this

A claim we couldn’t find anywhere in your copy. It may be true. You just didn’t say it.

Matches you

Word for word, or close enough. Shown as a count with a small sample.

Couldn’t check

We couldn’t compare this time, always with a short reason: the store blocked us, or we couldn’t finish checking the product (we try again the next day).

When your product page doesn’t load, we still check the product against your description, tags and product details, and the finding says so: “Checked on your description only.” Treat those findings with a little more caution.

What “serious” means

Inside each type, findings are sorted by severity. These are the definitions our AI reviewers are given.

Severity	What it covers
Serious	Health, medical or efficacy claims. Safety or suitability, including age ranges and choking warnings. Certifications. Free-from claims, like “fragrance-free.”
Ingredients & specs	Ingredients, concentrations, sizes, counts, origin and dosage.
Minor	Texture and packaging.

A “fragrance-free” claim is Serious even though it’s about an ingredient, because it’s a free-from claim. “Made in the USA” is Ingredients & specs.

When we re-check

Installed stores are re-checked once a day (on Free, the 5 products you updated most recently), and again after you edit a product. When you edit a product, we hold alert emails about it for up to 72 hours, so you aren’t emailed about Shopify catching up to your own change. The re-check itself still runs, and anything still true after the hold is sent then.

If you asked for alerts after a free scan, we re-check those 5 products once a day.

The limits

• One snapshot. Each check is what Shopify’s catalog said at that moment. Summaries change, which is why we re-check daily.
• Top results only, for the study. Our public study read the top 10 results for 20 searches. Your report covers your own products instead.
• Listings change. Yours and Shopify’s. A finding can disappear on its own.
• What agents are told, not what they say. We read Shopify’s catalog. Each agent may use that data a little differently, and we can’t see inside them.
• AI reviewers can be wrong. That’s why every finding quotes both texts in full, so you can check our work.
• We can’t edit Shopify’s summary. Nobody outside Shopify can. Fix notes only suggest changes to your own copy.