Ecommerce schema markup: what each type earns you

Two results, same product, same price

One result in Google is a blue link and a sentence. The one under it shows a price, the words "In stock", and a little trail reading Home › Lighting › Floor lamps. Same product, same price, same shipping.

The second store did not rank higher because of that markup. It got clicked more because of it. That distinction is the whole subject, and getting it wrong in either direction wastes effort: people either dismiss schema as cosmetic, or they install four plugins expecting a ranking boost that structured data has never provided.

Ecommerce schema markup is a block of JSON in the page source that states the facts of the page in a format machines read without guessing. Price. Availability. Where the page sits in your catalog. It changes what your result can look like, and increasingly what an AI assistant can repeat about your product without visiting your site.

The four types a store actually needs

Ignore the rest of the vocabulary. For a store selling things, these are the four that do work.

Product — the page is about a specific product. Name, image, description, brand, SKU, GTIN if you have one. This is the container the rest hangs from, and without it nothing else on a product page is eligible for anything.

Offer — the commercial facts: price, currency, availability, and sometimes the shipping and return details. Offer is why a result can show "$48.00 · In stock". It is nested inside Product and it is the part that most often goes stale, because it has to change when a product sells out.

BreadcrumbList — states the path from your homepage to this page. It replaces the raw URL in the search result with a readable trail, which reads as more trustworthy, and it tells a crawler how your catalog is organized. It is the cheapest of the four to add and the most commonly missing.

FAQPage — marks up questions and answers that are visibly on the page. Read the honest section below before you spend a day on this one.

What each type can actually earn

Two honest notes on that table. FAQPage used to reliably produce the expandable question list under a result; Google narrowed those rich results in 2023 so that, in practice, they show for a small set of authoritative government and health sites and almost nothing else. Marking up your FAQ is still reasonable — answer engines read it, and it costs nothing if the questions are already on the page — but do not plan around the rich result returning.

And none of these are guarantees. Valid markup makes a page eligible for a richer result. Google decides per query whether to show one, and it will often decide not to.

Why invented review markup is the one real risk

Most SEO mistakes cost you a slow decline. This one can cost you a manual action, which is a different category of problem: a human reviewer at Google flags the site, rich results disappear across it, and getting them back means fixing the markup and filing a reconsideration request.

What triggers it is structured data that does not match what is on the page. The specific versions that get stores in trouble:

• Ratings nobody left. A "4.8 from 126 reviews" block on a product that has no reviews visible anywhere on the page.
• Reviews you wrote about yourself. Google's own guidance excludes review content the business creates about itself or collects from a related party.
• Site-wide ratings pasted onto every product. One store rating duplicated across a thousand product pages, as though each item had been reviewed individually.
• Prices that do not match. Markup saying $19 to win the click while the page charges $29.
• "In stock" on sold-out items. The most common accidental version, and the most annoying for the searcher who clicks.

The rule is simple enough to hold in your head: markup only what a visitor can see on the page, and keep it true. If you have three real reviews, mark up three real reviews. Three is not embarrassing. A fabricated 4.8 is a liability that sits in your page source permanently, and the pressure to add one is exactly why the first sales from your own customers are worth asking for — see getting your first ten customers.

What mAutomate emits, and what it does not

Worth being specific, because "has SEO features" is the vaguest claim in this industry.

Every mAutomate storefront emits, on every product page, Product markup with the product name, description and image, and a nested Offer carrying price, currency and whether the item is in or out of stock, taken from live inventory rather than typed in. Sitewide it emits OnlineStore markup naming the store and its canonical URL. Each store also gets its own sitemap.xml and robots.txt on its own domain, a canonical tag, and the OpenGraph tags that decide how a link looks when someone pastes it into a chat.

What it does not emit is review or rating markup. That is deliberate — a platform cannot generate reviews that exist, and generating ones that do not is the risk described above.

So if you want stars in your results, you need real reviews collected from real buyers, and markup that reflects exactly those. BreadcrumbList, likewise, is not part of what the engine currently outputs on its own; if your category trail matters to you, that is one to add and verify.

How to check yours in ten minutes

1. Open your own product page source and search for application/ld+json. If there is nothing, you have no structured data at all and that is the finding.
2. Run the URL through Google's Rich Results Test. It tells you which types were detected and what is invalid. Warnings about missing optional fields are usually fine to leave; errors are not.
3. Check a sold-out product. This is the test almost nobody runs. If the markup still says InStock, you have found the defect that matters most.
4. Open Search Console's enhancement reports after a couple of weeks. Google lists the product pages it parsed and the ones it rejected, which is the only report that tells you what was actually accepted rather than what is technically valid.

Do these in that order. There is no point validating markup you have not confirmed exists.

What schema will not do

It will not move you up the page. There is no ranking bonus for structured data, and a well-marked-up page on a weak site stays where it was. The on-page and link work in the ecommerce SEO guide is what changes position.

It will not force a rich result. Eligibility is not display, and the same page can show a price for one query and a plain link for another.

It will not survive a catalog change on its own. Markup generated from live data stays accurate; markup pasted into a template as text goes stale the first time you reprice something, and stale markup is worse than none.

And it will not make a thin page worth visiting. Marking up a product page with nothing unique on it produces a good-looking result that leads somewhere disappointing, which raises your click-through rate and your bounce rate at the same time. Fix the page first — the order is in the product page SEO checklist — and let the markup describe something worth clicking.