Structured data for the page types that earn rich results, with nothing filled in on your behalf.
Pick a page type, fill in what you can confirm, and take the JSON-LD away. Anything you leave blank stays out of the output.
Everything here is worked out in your browser. Nothing you type is sent anywhere, and there is no account to make.
One per line.
Recommended rather than required. Google needs the price; without a currency the offer still validates but a merchant listing will not.
All ten values schema.org defines. Left blank on purpose unless you know. Claiming stock you have not checked is a claim to Google, and it is one of the fastest routes to a manual action.
Only if the rating is visible on the page itself. Both this and the count are needed.
One per line. Social and directory profiles that belong to the same entity.
Questions
Breadcrumb trail
Result
Fill in the fields on the left. Nothing is generated until you do, and nothing is filled in on your behalf.
What comes out
How to use it
The fields change with it. A product asks for a price, an article asks for an author, and the breadcrumb asks for a trail.
Structured data describes what a visitor can see. Markup that describes something absent from the page is the most common cause of a penalty.
One block per page, describing that page. The same block on every page describes the wrong thing on all but one of them.
Schema markup is a block of structured data that tells a search engine what the things on a page are: this string is a price, this one is a brand, this number is a rating out of five. Without it a crawler is reading a layout and inferring.
JSON-LD is the format to use. It sits in a script tag in the head, it is separate from the visible markup, and it does not break when someone redesigns the template.
What it earns is eligibility for richer results and a clearer link between your pages and the entity behind them. What it is not is a ranking factor on its own, and any tool implying otherwise is selling something.
Most generators fill the gaps. Availability defaults to InStock, a rating appears because the field existed, a currency is assumed from your language. Every one of those is a claim made to a search engine on your behalf.
Two of them were measured in this repository and removed. An unknown stock value was being written out as InStock, and an aggregate rating was being produced with no reviews behind it. Neither is only wrong data: both are the kind of claim that earns a manual action.
So nothing is filled in here. Availability offers all ten values schema.org defines and preselects none of them. A rating appears only when you enter both the value and the count. A blank field stays out of the output.
The tool
Product, article, FAQ, organization and breadcrumb, each with the fields that type uses.
Availability, currency and rating start empty. A guess in structured data is a claim, and it is made in your name.
Schema.org defines ten and this tool lists ten. An earlier version carried four, and the missing six were silently dropped instead of being offered.
An average with no count, or a count with no average, produces no rating block rather than half of one.
A product with no currency still validates, but a merchant listing will not accept it. The tool says so rather than letting you find out later.
A single block in the head, independent of your template, which is the format search engines ask for.
Reading the result
We look at your search visibility across Google and the answer engines, and tell you what is costing you orders. No charge for the first look.
In practice
Name, image, price, currency and a stock value you have checked. That is the set a merchant listing wants.
The questions you already answer on the page, in a form a machine can read question by question.
An organization block with the profiles that belong to the same entity helps a search engine connect the pieces of you it finds elsewhere.
A trail a visitor can see and a crawler cannot is a common gap on category pages, and it takes one block to close.
Questions
Not by itself. It makes a page eligible for richer results and helps an engine understand what the page is about, which are worth having. Treating it as a ranking lever leads to marking up pages that do not need it.
JSON-LD. It sits in one block in the head rather than being woven through the visible markup, which means a template change cannot quietly break it.
No, and this is where sites get into real trouble. A rating in the markup that a visitor cannot see on the page is a claim to a search engine, and it is one of the fastest routes to a manual action.
The one that is true right now. If you do not know, leave it blank: the markup validates without it, and an availability you have not checked is worse than one you did not state.
In the head of the page it describes, inside a script tag with the type set to application slash ld plus json. One block per page.
No. The JSON-LD is built in your browser and stays there until you copy it.
The rest of the set
Structured data tells a parser what a page is about. It earns rich results in classic search, while its effect on being cited by an assistant is not established, so plan it as SEO work rather than as an AI lever.
Where this comes up
