Short answer: the built-in layer covers the floor, not the game. Whether you should add more depends on which AI surface you are optimizing for, because there are two different data paths and schema only controls one of them.

Path one is the page itself. Crawlers like Googlebot read your HTML and its JSON-LD. Here, audits of standard Shopify themes find the default output is minimal Product markup: title, price, availability, and not much else.

That is valid schema. It is also the least any store on the platform ships, so it differentiates nothing.

Path two is the feed. AI shopping agents on Shopify surfaces read structured product data through Shopify Catalog, and Google's AI shopping features read Merchant Center data, not your page markup. Shopify's own guide to Google AI shopping is explicit that these experiences rely on complete, well-structured Merchant Center product data.

So the honest audit has two halves.

On the page, the gaps worth closing are specific. Brand, SKU and GTIN inside Product markup, since identifiers are how engines match and deduplicate you across the web. AggregateRating and review markup, because reviews are a signal both agents and rankings weigh; many review apps inject this, so verify with the Rich Results Test rather than assuming.

Add shipping and return policy markup on offers, since assistants quote policies before purchase. Add Organization markup with your name, logo, sameAs links and support contact, because entity clarity is how systems connect your products, your brand and what the web says about you. Add BreadcrumbList, and CollectionPage with ItemList on category pages.

Skip decorative schema. FAQ markup on invented questions, HowTo on pages with no steps, and Review markup without visible reviews add risk and no reach.

In the feed, "more schema" translates to attribute completeness. Fill titles, descriptions, variants, GTINs and availability the way agents query them, and keep Shopify's AI optimization guidance in mind: specifications, comparison information and structured attributes are what it asks for beyond the basics.

One more reason the page layer still matters even in a feed-first world. Classic crawling and indexing did not go away; AI answers are assembled on top of it, and how Google processes ecommerce pages still determines what is available to be cited. Schema will not rank a weak page, but incomplete schema silently removes a strong page from rich results and comparison surfaces.

A practical order of work for a normal store. First, validate what your theme and apps already output, because duplication and conflicts are common. Second, close the identifier gap: brand and GTIN everywhere. Third, get ratings markup rendering. Fourth, add Organization and policy markup once, sitewide. Fifth, move to the feed and treat attribute completeness as the ongoing job.

The straight verdict: built-in structured data makes you eligible. It does not make you preferred. Preference comes from identifiers, reviews, policies and attribute depth, and most of that lives in your product data before it ever becomes markup.