By Steve Merrill · September 5, 2026
How do you audit your Shopify structured data for AI search?
Open three product pages, one collection page, and your homepage, then check the schema on each with the Google Rich Results Test and the Schema.org validator. You are looking for one clean block per type, real values in every required field, and no duplicates. That is the whole audit in one sentence. The rest is detail.
We put out a checklist version of this a while back. This is the walkthrough. If you want the short reference to print and tape to your monitor, grab our 2026 structured data audit checklist and keep it open next to this post.
Here is why this matters now. AI engines read your schema to understand what you sell before they decide whether to cite you. Bad schema means you are invisible to the thing sending buyers. Good schema means you are in the running. Let's go page by page.
What should you check on product pages?
Every product page needs one Product block with name, image, description, sku, brand, and an Offer that includes price, priceCurrency, and availability. Miss any of those and the page reads as incomplete to a machine.
Right click a product page and view source. Search the page for "application/ld+json" and read what comes back. You want to see one Product block. Not two. Not three.
Check the availability field on an out-of-stock item and a sale item. This is where stores get burned. The visible page says sold out, the schema still says InStock, and now you are lying to the parser without knowing it. Shopify's own structured data documentation walks through where these fields come from in most themes.
I have seen this exact pattern in dozens of audits. The store looks fine to a human and reads as broken to a bot.
What about collection pages?
Collection pages should carry CollectionPage or ItemList markup that names the products sitting in the grid. Most Shopify themes ship these pages with almost nothing, so a machine sees a page of links and no meaning.
This is the most ignored page type on the whole store. People pour effort into product schema and leave the category pages blank. AI engines love category pages because they answer "best X for Y" questions directly.
Check that the ItemList actually lists the products shown, in order. An empty or stale list is worse than none. Fix the template once and every collection inherits it.
How do you check homepage and Organization schema?
Your homepage needs exactly one Organization block with a stable name, a url, a logo, and sameAs links pointing to your real social and business profiles. This block is how AI engines connect your store to your brand as an entity.
The word that matters here is stable. Your Organization name has to be spelled the same way everywhere. Same on the homepage, same in your product brand field, same on your social profiles.
Pull up your homepage source and confirm there is one Organization block, not one from the theme and another from an app. The sameAs array is the part most stores skip. Add your Instagram, your TikTok, your LinkedIn, your real business listings. That is how you become a known entity instead of a random URL.
What do you look for in FAQ and HowTo blocks?
Any FAQPage or HowTo markup on a page must match the visible text word for word, and no page should carry two of the same type. Schema that describes questions a visitor cannot see on the page is a violation, and duplicates cancel each other out.
Here is where apps cause trouble. You install a review app, an FAQ app, and a theme that all decide to add their own blocks. Now one product page has two FAQPage blocks with different content. Google's FAQPage guidelines are clear that the markup has to reflect content the user actually sees.
View source, search for "FAQPage" and "HowTo," and count. One each, max, per page. Delete the rest at the app or theme level.
How do you test schema with Google and Schema.org?
Run each page type through the Google Rich Results Test and the Schema.org validator, then fix every error and clear the warnings that apply. The Rich Results Test tells you what Google can render. The Schema.org validator tells you whether your markup is technically valid at all.
Use both. They catch different things. Rich Results is strict about what earns a visual result. Schema.org is strict about whether your JSON even makes sense.
Test one page per template. Shopify builds from shared templates, so one passing product page usually means the whole template passes. Then spot check your weird edge cases. Sold out. On sale. Bundle. Those break in ways the default product never does.
What are the new 2026 checks?
Three checks matter more this year than they did last year: freshness through dateModified, entity consistency across pages, and zero duplicate FAQPage or HowTo blocks. AI engines now weigh how recent and how consistent your data is, not just whether it validates.
Freshness first. Your dateModified should move when you actually edit a page. A field frozen at 2023 tells an AI engine the page is stale, even if the content is current. Search Engine Land has covered how recency signals feed into AI answer selection.
Entity consistency second. Same brand name, same spelling, same casing, everywhere it appears. Mixed spellings split your entity in two and weaken both halves.
Duplicates third. This is the one that quietly kills stores. Run the count check on FAQPage and HowTo across your main templates. If a theme update or a new app added a second block, you will find it here. Not glamorous. Just the thing that moves the needle.
Frequently asked questions
How often should I audit my Shopify structured data?
Quarterly, plus any time you change themes or apps. Run a full audit once a quarter, and run a quick spot check any time you switch themes, add an app, or edit your product template. Theme and app changes are the most common way schema breaks without anyone noticing.
Does structured data guarantee my products show up in AI search?
No. Clean structured data makes your pages readable and eligible. It does not force a citation. AI engines still weigh trust, freshness, and how well your content answers the question.
What is the biggest structured data mistake Shopify stores make?
Duplicate schema. Two apps and a theme all add their own Product or FAQPage block on the same page, and the conflicting data confuses parsers. One clean block per type per page beats three fighting each other.
Do I need to test every single product page?
No. Test one page per template. Shopify builds product pages from a shared template, so if one product page passes, the rest of that template usually pass too. Spot check a few edge cases like out-of-stock and sale items.
Want us to run the audit for you?
If you would rather have someone go through your store page by page and hand you a fixed set of schema, that is what we do. See how WRKNG Digital gets Shopify stores cited in AI search.
