Product data checks
A product data audit checks whether a product page gives AI assistants and shoppers what they need to understand and compare it: key specifications, use conditions and limits, and claims backed by sources.
Last updated September 14, 2026
When a shopper asks an AI assistant what to buy, the assistant can only work with the product facts it can read. A product data audit looks at one product page, or a catalog, and answers three questions: is the product clearly described, are its claims backed by evidence, and what should change first.
It is a good fit for brands and ecommerce merchants whose products are compared on specifications, use conditions or evidence: electronics, outdoor gear, beauty, home goods, software and similar categories. It is less useful for pages with no product or service to compare, or for teams that only want classic keyword-rank tracking.
The audit does not tell you why a particular AI assistant did or did not recommend a product. Assistants do not publish those reasons. It tells you what your own product information is missing, and gives you a way to check it again after you change it.
Hasmord checks product information across six dimensions. They are Hasmord's framework for reviewing product information, not the ranking factors of any AI assistant.
Checks are organized by category, because different products need different facts. Hasmord currently supports 41 categories in eight groups: electronics and tech; fashion and accessories; home and living; health, beauty and lifestyle; automotive; kids, baby and pets; food and consumables; and software and digital products.
Not every gap is the same kind of problem, so the report keeps them apart:
| Finding | What it means | Typical next step |
|---|---|---|
| Missing | A fact buyers compare on is not on the page at all. | Add the fact, if you know it. |
| Conflicting | The page, the source or another listing say different things. | Decide which value is correct and fix the others. |
| Unconfirmed | A claim is stated but no matching source supports it yet. | Add a source, narrow the claim, or remove it. |
How claims are matched to sources is covered in product claims and evidence checks.
Illustrative example. The product, fields and sources below are invented to show the format of a report. They are not a customer result.
Product: Example trail running shoe, model TR-2 (category: footwear).
| Item | Finding | Suggested change |
|---|---|---|
| Sizing | Missing — width options are not listed. | Add the available widths and a sizing note. |
| Weight | Conflicting — the page says 280 g; the spec sheet says 295 g for size 9. | Confirm the value and state the size it applies to. |
| "Waterproof" | Unconfirmed — no test or membrane specification is linked. | Link the source, or change the claim to what is supported. |
| Best for | Missing — no statement of terrain or distance it suits, or does not suit. | Add a short "best for / not ideal for" section. |
| Returns | Present — return window and conditions are linked from the page. | No change. |
The free check on the homepage scores one product page with no signup. Plan limits for larger catalogs are on the pricing page.
Related: the six readiness signals in detail · product data audit checklist · why isn't AI recommending my product? · product data checks for Shopify · ChatGPT shopping checklist.
A review of a product page, or a catalog, for missing, conflicting and unconfirmed information, so the product is easier for AI assistants and shoppers to understand and compare.
No. AI assistants do not publish why a product was or was not included. The audit shows gaps in your own product information, which you can fix and check again.
Accessible public product pages, including Shopify, WooCommerce and BigCommerce storefronts. You can also import products by CSV.
No. A person on your team reviews and approves each change, and approved changes can be rolled back.