Structured data using Schema.org vocabulary to describe things such as organizations, products, articles, and services.
Schema markup uses the Schema.org vocabulary to describe things and their properties. It can identify that a page concerns a product, an article or an organization and label relevant information about it. The vocabulary gives producers and consumers a shared way to name those concepts.
Keep three layers separate: the facts being described, the vocabulary used to describe them and the format carrying the data. Schema.org supplies vocabulary; JSON-LD is one way to express it. A rich result is a possible search presentation, not another name for the markup itself.
Start with what the page actually represents. Choosing a type because its name sounds commercially attractive can create inaccurate data. The presence of a Schema.org type also does not mean Google offers a special search feature for it. Check the search feature’s own documentation when enhanced appearance is the goal.
Takeaway: Use schema to describe real page information consistently. Select an appropriate type and maintain its facts rather than add unrelated markup in the hope of receiving special search treatment.
Example
A business publishes an article about a completed project. It can describe the article and its author without pretending the article is a product for sale. If the business also has a genuine product page, that page needs its own accurate product information. The vocabulary choice should follow the content’s role.
What to check
Validate the vocabulary and syntax, then review the facts against the visible page. A validator can flag structural problems but cannot establish that an invented rating or inaccurate author claim is true. Keep maintenance responsibility clear when a plugin generates the data automatically.
For more detail, see Getting started with Schema.org.