Google Merchant Center Audit: A Step by Step Guide for Shopify

Quick answer: A Merchant Center audit checks three layers: whether your products carry the attributes Google requires, whether they carry the extra detail shoppers actually search with, and whether your feed still matches your live product pages. Diagnostics only covers the first. The second layer is where most of your missing impressions live.

Most merchants audit Merchant Center the same way. Open Diagnostics, see no red errors, close the tab. That check answers one question, whether anything is broken badly enough to be rejected. It does not answer the more expensive question, which is how much of your catalog is technically approved and still invisible.

We covered why that happens in Why Your Fashion Products Don't Show on Google Shopping. This post is the practical follow up, focused on Google: how to run the audit, in what order, and what to do with the result.

The three layers a Merchant Center audit has to cover

Eligibility. The attributes that trigger disapprovals. Miss one and the product leaves Shopping ads and free listings entirely. For apparel that means color, size, gender, and age group, plus brand, a valid identifier, price, availability, and a working image link. These failures are loud, and Diagnostics reports them.

Enrichment. The attributes nobody requires and everybody competes on. Material, fit, silhouette, sleeve length, neckline, pattern, occasion. Nothing breaks when they are empty. Your product simply cannot appear for a search like "black linen midi dress with sleeves," because there is no linen, no midi, and no sleeve anywhere in your data. These failures are silent, and they are why approved products earn nothing.

Consistency. Price, availability, and image have to match between your feed and the landing page. This layer breaks after the fact, usually the week you launch a sale.

An audit that only reads Diagnostics will tell you the catalog is healthy while most of it sits unseen.

The audit, step by step

1. Read Diagnostics first, then stop trusting it

Open Merchant Center Diagnostics and group what you see by error type rather than by product. A hundred item level errors usually trace back to two or three root causes.

Fix account level issues before anything else, since those can suppress entire campaigns while you are busy editing attributes. Then handle feed level issues, which affect whole batches, and only then item level ones.

Then note the ceiling. Diagnostics reports what is wrong. It does not report what is thin. A product with a three word title, no material, and no pattern passes every check on that screen.

2. Export your catalog and count gaps at variant level

Export your Shopify products as CSV with variants included. This is the baseline everything else runs on.

Check the required apparel attributes on every variant, not every product. This is where fashion catalogs quietly fall apart. A store with 400 styles in six sizes and four colors is auditing close to 10,000 rows, and a size value present on eleven variants and blank on the twelfth is enough to pull that twelfth listing out of Shopping.

Count blanks per attribute. You want a number, not an impression. "Size missing on 312 variants" is actionable. "Sizing looks inconsistent" is not.

3. Check publishing status before you check anything else

This one catches people out because it does not look like a data problem. A product that is not published to the online store has no landing page, and without a landing page Google cannot list it at all. Every attribute on that product is irrelevant until it is published.

Filter for unpublished, draft, and archived products first. Then check availability, because Google drops out of stock products until they are back, and a shelf of permanently out of stock items quietly drags your catalog down.

4. Get brand and identifiers in proportion

Brand should be on every product. It is cheap to fill and both channels use it.

Barcodes are more nuanced than most guides suggest. If your products genuinely have no manufacturer GTIN, which is normal for own label fashion, you are still fully eligible. Declare the absence properly rather than leaving the field ambiguous, and put your effort into titles and attributes instead, since that is what Google will match on.

If some of your products do carry a manufacturer GTIN, add it everywhere it applies. That unlocks price benchmarks and stronger product matching, but treat it as an upgrade rather than as the blocker it is often described to be.

5. Check your Google product category and product type mapping

If your Shopify collections map loosely to Google's taxonomy, your products end up competing in the wrong neighborhood. That never shows as an error. It shows as poor performance.

Build an explicit lookup from your collection structure to the closest Google taxonomy node, rather than letting an automatic mapping guess. Fashion taxonomy is deep, and the difference between a generic apparel node and the correct garment node is the difference between broad, expensive matching and precise matching.

Category also gates everything else. Apparel specific requirements only apply once the product is classified as apparel, so an unset category quietly disables the checks that should have been protecting your listings. Set the Shopify category and the rest becomes derivable.

6. Score your enrichment, not just your completeness

Pick the attributes that matter for your category and measure coverage the same way you measured the required ones. Three groups are worth separating.

Reach attributes. Material and pattern are the big two. Material is the single most commonly missing apparel attribute, and it is what powers every "linen shirt" or "wool coat" search and filter. Pattern unlocks the print based queries, striped, floral, plaid.

Market attributes. Size system and size type. Without a declared size system, a size value is ambiguous the moment you sell across borders, and size type is what lets petite, plus, tall, and maternity shoppers find the cut they need.

Findability attributes. Fit, neckline, occasion, sleeve length, style. These are almost never captured structurally in Shopify, which is exactly why they are an opportunity. They are standard onsite filters and they carry real search intent, and a catalog that structures them competes for queries its rivals cannot reach.

Required attributes keep you eligible. These decide how many different searches you are eligible for. A product carrying eight descriptive attributes can match dozens of long tail queries that a product carrying two cannot reach at any bid.

7. Audit titles against the attributes you just counted

Read fifty titles, starting with your highest revenue products. Count how many name a color, a cut, and a key feature.

"Floral Dress" competes for almost nothing. "Green Floral Print Midi Wrap Dress with Short Sleeves" competes for a great deal. A workable structure is brand, then product type, then the attributes that distinguish it.

The same detail belongs in your description and structured data, which is what free listings and AI shopping assistants read. Those systems match on words. If a detail was never written down, nothing can surface it.

8. Spot check feed against live pages

Sample twenty to fifty products and verify that price, availability, and image in the feed match what a shopper sees on the product page. Price mismatch causes disapproval on its own, no matter how clean your attributes are, and it usually appears right after a promotion rather than during setup.

9. Rank every gap by number of products affected

This is the step most audits skip, and it decides whether the work is worth doing. A missing price on 2,000 products and an invalid URL on 50 are not the same task. Sort every gap by how many products it touches, then by effort, and work the top of that list.

Why the manual version stops working

Everything above is doable. It is also, for a real catalog, several hours of spreadsheet work that has to be repeated every quarter and after every collection drop.

The variant math is what breaks it. A mid sized fashion store is auditing tens of thousands of rows against roughly twenty attributes per row. Doing that by hand once is a project. Doing it four times a year is not going to happen, which is why most catalogs are audited once, at setup, and then never again.

That is the gap we built a tool for.

Run the same audit in minutes, free

SixFit Catalog Checkup is a free Shopify app that runs this audit against your live catalog. It reads your products and checks each one against the fields Google Shopping uses to decide what shows and what ranks.

What you get:

• Two scores out of 100, completeness and enrichment, measured separately. Layer one and layer two apart, because a catalog scoring 76 on completeness and 24 on enrichment has a very different problem from the reverse, and a single blended number would hide both.
• Channel eligibility, counted. How many of your products qualify for Google free listings and Google Shopping ads, alongside Meta catalog and Meta variant ready, since all four draw on the same product data. A product counts as eligible only once it carries every field that channel requires, so the number is usually lower than merchants expect.
• Every gap, ranked by how many products it affects, and sorted by tier: required fields that block listing, recommended fields that lift ranking, nice to have fields, and findability attributes like fit, neckline, occasion, and sleeve length. Step nine, done for you.
• A plain explanation of why each field matters, so the report is something you can hand to whoever owns the catalog without translating it first.
• The context that changes your reading of the score, including out of stock products, drafts excluded from scoring, and whether your barcode situation actually counts against you.
• A shareable report you can save, print, or send on.

It reads products and collections only. It does not write to your store and does not touch orders or customer data.

Install SixFit Catalog Checkup, free

What to do with the report

Fix at the source, not in the feed. A feed tool can only send data that already exists. If a garment's fabric, fit, or true color was never captured on the product, the feed has nothing to move. Close the gap on the product and every channel downstream inherits it.

Eligibility first, enrichment second, but do not stop at eligibility. Required attributes protect the listings you already have. Enrichment grows them, and that is where the impressions you are not currently earning actually live.

Measure it. Hold back a group of products, fix the rest, and compare impressions and revenue after a few weeks. A simple holdout tells you whether the work moved anything, which beats guessing and makes the next round easy to justify.

Meta runs on the same product data but fails on a different set of fields, including some Google never asks for. If you advertise on both, see our companion guide: How to Audit Your Shopify Catalog for Meta Ads.

One more thing if you sell apparel. The same garment detail that makes a product legible to Google is what tells a shopper whether the item will fit her. An analysis of one million fashion returns found that 67 percent were driven by fit and sizing issues. Better data gets the product seen. The right data gets it bought and kept.

Sources

1. Google Merchant Center Help, "Fixing Merchant Center disapprovals for product data quality violations." On products with data quality warnings serving with limited performance, lower impressions, and lower clicks. support.google.com/merchants/answer/13693497
2. Google Merchant Center apparel requirements. Required attributes for clothing, including color, size, gender, and age group.
3. DataFeedWatch analysis, published via Search Engine Land (2023). On the share of Google Shopping products disapproved for critical errors.
4. Inriver, "Product data requirements for Google Shopping: A complete reference" (2026). On a single missing attribute or price mismatch removing a product from Google Shopping.
5. Verde Media, "Google Shopping Feed Quality: Impact on Ad Performance, CTR, and Revenue." On how missing color, size, and material attributes limit which searches a product can appear for.
6. PRIME AI, analysis of one million fashion returns. Finding that 67 percent of returns were driven by fit and sizing issues.

See what Google can actually read in your catalog

Most stores are surprised by how much of their catalog is sitting invisible right now. You can find out in a few minutes, for free.

Install SixFit Catalog Checkup on the Shopify App Store and get your score, your gaps, and your fix list in one report.