Fix Google Merchant Feed Errors: A Step-by-Step Validation Playbook

Merchant Center feed errors — disapproved items, invalid identifiers, missing required attributes — are almost always a symptom of upstream data quality problems, not one-off typos. When a supplier delivers a range of MRO fittings or CPG SKUs with blank GTINs, vague titles, and inconsistent availability values, those problems appear as a wall of diagnostics that grows back every sync no matter how many rows you fix inside the UI.

Claro resolves this at the source. It identifies which records have bad or missing identifiers, enriches attributes against trusted supplier and catalog data, validates every value against channel rules, and writes the corrected records back into your PIM or ERP — so the next feed export is already clean before it reaches Merchant Center. This playbook walks you through the end-to-end process: from pulling a diagnostics export to confirming zero blocking errors on reupload, with or without automated tooling.

Steps to fix Google Merchant feed errors

1. 1
   Pull a fresh diagnostics export

   In Merchant Center, open Products then Diagnostics and export the item-level issues. Sort by issue severity: “Disapproved” items are blocked from serving; “Demoted” items still serve but rank lower. Group errors by type — for example, missing gtin, invalid price, image too small — so you fix root causes once instead of editing rows one at a time. A CPG feed with 400 disapproved items usually has three or four underlying issues, not 400 unique ones.

2. 2
   Confirm required attributes are present and populated

   Every offer needs id, title, description, link, image_link, availability, and price. For most branded products Google also expects a valid gtin or a brand + mpn pair. Check for blanks AND for placeholder junk like “N/A”, “0”, or “TBD” that passes a not-null check but fails Google’s validation. A furniture catalog missing gtin on assembled bundles, or an MRO feed with empty mpn on private-label fittings, will see disapprovals cluster at this step.

3. 3
   Validate identifiers, not just their presence

   A populated gtin is not the same as a correct one. Run the GTIN values through check-digit and length validation, and confirm brand/mpn pairs actually resolve to real products. Inconsistent or invented identifiers are a top cause of “Invalid value” and “Incorrect identifier” errors in CPG and industrial feeds. Claro validates each identifier against source records and flags values that cannot be verified so you never silently ship a fabricated code.

4. 4
   Fix formatting, encoding, and price and currency issues

   Watch for the silent killers: a wrong delimiter splitting columns, UTF-8 encoding problems turning “Ø” into garbage, prices missing an ISO currency suffix (19.99 USD), or availability values that do not match Google’s controlled vocabulary (in_stock, out_of_stock, preorder, backorder). Normalize these in the source file — your PIM export or ERP feed template — not by hand inside Merchant Center, or the next sync will reintroduce every one.

5. 5
   Map products to the right Google product category

   Many “Missing value [google_product_category]” errors and miscategorization flags come from feeds that ship without explicit categories or with categories that do not match the product. Map each item to a valid Google taxonomy node so an industrial coupling does not land under Home and Garden, and so variant-heavy apparel lines reach the right leaf node. The Google Product Category Finder can speed this step.

6. 6
   Reupload, recrawl, and confirm zero blocking errors

   Push the corrected file, trigger a fetch, and recheck Diagnostics. Confirm disapproved counts dropped to zero and that no new errors appeared from your edits. Schedule a recurring validation so the next supplier update or price change does not silently reintroduce the same failures. If you are using Claro’s write-back, the canonical record in your PIM or ERP is already updated, so future exports inherit the clean state automatically.

Messy feed vs. trusted feed

The difference between a feed that generates daily disapprovals and one that ingests cleanly comes down to data state before export, not post-upload patching.

Common pitfalls

Other recurring traps:

A clean Merchant Center feed is also the foundation for AI discovery. The same complete, verifiable attributes that satisfy Google’s diagnostics — accurate titles, valid identifiers, correct categories, complete specs — are what large language models cite when shoppers ask for product recommendations. Feeds that are validated, enriched with provenance, and written back cleanly into your existing PIM or ERP stay clean across every future export without manual re-intervention.

Related

FAQ