Google Shopping Product Feed Quality: A Practical Checklist

In Shopping ads there is no keyword list to tune. Google reads your product feed and decides which searches your products fit. If the feed is thin, wrong or out of date, no bidding strategy repairs it. The feed is the targeting.

This checklist helps you audit it in a sensible order and decide which fixes are worth the time first.

Why feed work gets skipped

Feeds are usually set up once by whoever connected the store, then left alone. Problems stay invisible because ads still run. Owners look at bids and budgets, which are easy to change, rather than at titles and identifiers, which take work.

The result is predictable. Products that are mislabelled, missing identifiers or priced differently on the page than in the feed may be limited or not shown at all. Merchant Center reports issues, but only if someone reads them.

The audit model: four layers

Audit in this order, because each layer depends on the one before it.

1. 1

   Eligibility

   Are products approved and showing, or limited by issues?

   1
2. 2

   Accuracy

   Do price, availability and links match the product page?

   2
3. 3

   Identity

   Are product identifiers and variants described correctly?

   3
4. 4

   Relevance

   Do titles, descriptions and categories match how buyers search?

   4

Do not rewrite titles while products are still disapproved for a price or availability mismatch.

1. Eligibility

Open the diagnostics or issues view in Merchant Center. Group the problems by type rather than by product, so one fix clears many items. Separate errors that stop products from showing from warnings that only reduce quality.

2. Accuracy

Price, availability and the landing page link must agree with what a buyer sees after clicking. Google's specification treats these as core attributes, so a mismatch is the first thing to remove. If prices change often, check how your feed updates and how often. The documentation describes update options; read the current page for your setup instead of assuming a schedule.

3. Identity

Identifiers help Google understand exactly which product you sell. The specification says that when you provide a GTIN it must have the correct length and check digit. Use the manufacturer's real code when one exists, and do not invent values to clear a warning. If your products have no manufacturer code, check the specification for what to provide instead.

Variants (size, colour) should be described consistently and grouped as the documentation explains. Inconsistent variant data is a common source of confusing listings.

4. Relevance

Now improve the words. The title is usually the most influential text, so put what a buyer would type first: product type, brand where relevant, and the attributes that separate it from similar items, such as size or material. Write descriptions for a human reader and keep them accurate. Do not stuff keywords or add promotional text the specification does not allow. Check the current limits and rules in the documentation rather than relying on remembered numbers.

• Title

  Leads with the product type and the attribute that matters.

• Description

  Accurate, readable, no repeated phrases.

• Price and availability

  Match the landing page exactly.

• Identifiers

  Real GTIN, brand and variant data where they exist.

• Category and images

  Correct product category and a clear main image.

Names and requirements can change, so confirm each in the current specification.

A worked illustrative example

Illustrative example, not real results. A feed has 1,200 products. Merchant Center shows 84 products with a missing identifier and 36 with a price mismatch.

• Affected: 84 + 36 = 120 products, or 10% of the feed.
• Clean: 1,200 − 120 = 1,080 products, or 90%.
• Price mismatches are 3% of the feed but can stop products showing, so they come before the 7% with missing identifiers.

Fix order follows impact on eligibility first, then volume. The percentages are only a way to prioritise; they say nothing about sales.

Trade-offs and when this changes

• Rewriting thousands of titles by hand is slow. Rules in your feed tool can handle patterns, while your best-selling items deserve manual attention.
• If you sell mostly your own unbranded products, identifier fixes matter less than titles and images.
• If your catalogue is small, a simple spreadsheet feed may be better than a complex tool.
• If products change price or stock daily, automation of updates matters more than wording.

Implementation steps

1. Export the current issues and group them by type.
2. Fix the accuracy problems at the source (the store), not only in the feed.
3. Correct identifiers and variants.
4. Rewrite titles for your top sellers first, then apply rules to the rest.
5. Re-check after the next feed update and compare the issue counts.

Decide whether Shopping fits your account at all with Google Search vs Shopping ads. The conversion tracking behind Shopping is in setting up Google Ads conversion actions, and what you can and cannot steer in automated campaigns is in Performance Max: what you can control.

Measurement limits

Feed quality is an input, not a result. Cleaner data can improve eligibility and matching, but it does not show that sales will rise, and you cannot tell from clicks alone which attribute made the difference. Change one group of fixes at a time where possible and read the effect against margin, not only revenue. If you want an independent review of the feed together with tracking and account structure, the full ads and tracking audit covers it.

Sources

• Google Merchant Center Help: product data specification
• Google Merchant API reference: product attributes