# Product feed disagreements: build a truth sheet before adding more traffic.

Reconcile feed attributes, visible product pages and checkout for representative variants so a catalogue campaign starts with consistent facts.

By AdzUp Editorial | 2026-11-20

## Treat product data as a shared business record.

Merchant Center’s current product data specification explains that missing, incorrect or conflicting product information can affect eligibility and display. Search Central also distinguishes product snippets from merchant listings for purchasable products. Neither a feed nor markup replaces an accurate shopping experience.

Give merchandising, website and media teams one small verification sheet. The goal is not to duplicate the entire catalogue manually. It is to expose where the same item is described differently across systems, especially when a promotion or inventory update is approaching.



## Sample the cases most likely to disagree.

Choose products with size or colour variants, sale prices, limited stock and delivery restrictions. Include one newly added item and one recently changed item. A sample made only of stable bestsellers can miss the synchronisation failures that matter during a launch.

Illustrative example: a textile retailer compares two sizes in the same fabric range. The page displays a range price, the feed lists a specific size and the cart applies another price. The discrepancy is a data-and-template question to resolve, not a reason to write a more persuasive ad.



## Record expected facts and observed facts.

For each sample, note a stable item identifier, variant, feed value, visible page value, structured-data value where used and checkout value. Add the observation time and whether the user was signed in. Prices and availability are time-sensitive; a screenshot without timing can hide the sequence of a change.

Use blank or unknown deliberately when a field has not been verified. Do not invent product identifiers, reviews or stock states to make a diagnostic look complete. Assign an owner to establish the source of truth for each disputed field before correcting downstream representations.



## Fix the update path as well as the item.

A one-off catalogue correction can be undone by the next import. Identify where the erroneous value originated and which process publishes it. Test whether the correction survives a normal catalogue update and whether a page cache delays the visible change.

Keep sample evidence for the corrected item and a neighbouring variant. If a shared template caused the problem, verify the wider affected set. If the issue was isolated, avoid a broad rewrite that creates unnecessary risk for otherwise accurate products.



## Approve promotion readiness with exceptions visible.

Create a short release decision: approved items, excluded items, unresolved issues and the person responsible for another check. If only part of the range is ready, a narrower campaign is more honest than presenting the full collection as available.

Review diagnostic feedback after the next update and track whether the same discrepancy returns. The acceptance condition is consistent, truthful product information across the actual buying route. It is not simply that the upload succeeded or the structured-data syntax parsed.

- Sample variants and recently changed items.
- Compare feed, page, markup and cart at a recorded time.
- Repair the publishing process that caused the mismatch.

## Sources

- [Merchant Center: Product data specification](https://support.google.com/merchants/answer/7052112)
- [Google Search Central: Product structured data](https://developers.google.com/search/docs/appearance/structured-data/product)
