Catch product data errors before a Merchant Center upload

The same shoes, a different color in the feed

Suppose you sell one shoe model in several colors and sizes. The manufacturer assigns a separate code to each version. Its spreadsheet lists navy beside the code for a navy pair in size 8. Your Merchant Center feed lists black beside that same code.

A program can put those two records together and show the difference to the employee preparing the feed. They can see the product code, both colors, and the supplier row the information came from. That lets them check and approve the correction without searching the whole spreadsheet again.

The code matters because you also sell those shoes in black. Comparing model names alone could match two different pairs and suggest changing a listing that was already correct. If the supplier and your store use different identifiers, first establish which supplier code belongs to which product in your catalog. Without that connection, a list of discrepancies could add work.

It is also worth checking the store's own record. If that says black, correct the data used to produce the feed. If the catalog already says navy and black appears only in the Merchant Center file, fix the export. That addresses the recurring error at its source instead of leaving someone to repair each new file.

What to do with an empty field

Missing information needs a different response. A program can recover a value left out when a spreadsheet was copied. If the supplier has not provided it anywhere, someone needs to ask. The color of another pair in the same model is not evidence for this listing.

Before investigating every blank cell, check whether the information is needed. Google's product data specification distinguishes required, conditional, and optional attributes. Requirements can depend on the product and target country, so color in this example should not become a requirement for your entire catalog (Google Merchant Center product data specification).

An optional field can remain empty. For a conditional requirement, establish whether it applies to this product and market. If a required value is missing, obtain it before submitting the item. A filled cell can also need attention when it contradicts the supplier's information or uses the wrong format.

Google identifies missing or incorrect variant attributes and differences between feed and website data as common issues. Missing or inaccurate information can lead to disapprovals, limited eligibility, or incorrect displays. Checking the file can catch errors before submission, though other requirements still determine approval (Google's product data requirements).

You can compare spreadsheets without AI

When suppliers send consistent spreadsheets and the product codes match, ordinary comparison may be enough to find differences and missing values. Review your existing import and export first. You may need to repair one connection between systems. Merchant Center's issue details can also help staff investigate problems Google reports after submission.

AI may help when employees search through manufacturer documents with different layouts. Instead of opening each file themselves, they could receive a suggested value with a reference to the page where it was found. They still need to establish that the document describes the right shoes. A useful suggestion makes that check easier and helps them finish the listing.

Before extending the checks to the whole catalog, show the provider listings your team knows, including correct ones. Let the employee who normally prepares them work from the proposed findings. If the program flags many correct records or leaves staff searching for the source of every suggestion, it needs more work. Use the next update to see how it handles new colors and sizes, and agree who will keep the checks current as product or market requirements change.

If similar corrections begin with purchasing data, our article on supplier price list checks covers that earlier step.

Start with a correction that keeps coming back

Syntalith builds automations that connect data across systems and documents. Start a conversation with an error your team fixes after each new supplier file. Explain where they find it and what they do next. That will help establish whether you need a feed comparison, help finding information in documents, or an export fix. See Syntalith pricing for cost information.