What Structured Data Actually Does
Schema markup is sold as a ranking trick and is not one. What it actually does is more useful, and it matters more now that answer engines are reading your pages too.
Structured data does not directly improve rankings. It makes a page machine-readable, which qualifies it for rich results in search and makes it easier for AI answer engines to quote accurately. The gain is click-through rate and citation accuracy, not position.
What It Is, and What It Isn't
Structured data is a block of machine-readable facts describing what a page is about — that this is an article, published on a date, by an author, or that this is a business with an address and opening hours. It states explicitly what a human reader infers from the layout.
Google has been clear that structured data is not a ranking factor in itself. Adding it to a weak page does not lift the page. What it does is qualify the page for enhanced presentation — review stars, FAQ dropdowns, event dates, breadcrumb trails — which changes how many people click a result that was already ranking.
That distinction matters commercially. Moving from position five to position three is hard. Doubling the click-through rate at position five by earning a richer result is often easier and produces a similar amount of traffic. Structured data is a presentation lever, not a position lever.
The second job has become more important recently. Answer engines reading your pages benefit from explicit facts in the same way search crawlers do — an address stated in markup is unambiguous in a way that an address in a footer image is not.
The Types Worth Implementing
There are hundreds of schema types and most sites need four or five. These are the ones that reliably earn something, in rough order of value for a service business.
Organization or LocalBusiness
Your name, address, phone, hours, logo and social profiles. This is the foundation: it helps Google resolve your business as a real entity rather than a string of text, which supports brand searches and local visibility. LocalBusiness and its subtypes add address and hours handling.
BreadcrumbList
Turns the URL line under your result into a readable path. Small, cheap, and it makes a result look more established than a bare URL. Worth adding to every page below the homepage.
FAQPage
Marks up genuine questions and answers on a page. Eligibility for FAQ rich results has narrowed considerably, but the markup still helps engines extract answers, which is increasingly where the value sits.
Article or BlogPosting
Declares author, publication date and publisher for editorial content. Feeds Google's understanding of freshness and authorship, and gives answer engines a clean citation with a date attached.
Product and Review, where genuine
Star ratings in results are the most visually powerful rich result available. They are also the most heavily policed. Only mark up reviews that are real, collected from actual customers, and visible on the page itself.
Connecting Your Entities
Isolated schema blocks describe pages. Connected ones describe a business. The difference is whether your markup uses stable identifiers so that the same entity referenced across many pages is understood as one thing rather than many.
Give each core entity a stable identifier — your organisation, your website, your author — and reference those identifiers from everywhere else rather than redeclaring the entity each time. An article whose author points at the same identifier across forty posts accumulates authorship signal on one person; forty separately-declared authors with the same name accumulate nothing.
There is one trap worth knowing. Search engines resolve these references within a single page's own markup, so a reference to an entity that is never declared on that page is a dangling pointer and gets dropped silently. If your article markup says its publisher is your organisation, the organisation block has to be present on that page too — which is why site-wide entities usually belong in a layout rather than on the homepage alone.
This is invisible in testing tools that validate blocks individually, which is why it survives so often. Validate the page as a whole, and check that everything you reference is actually there.
The Mistakes That Get Rich Results Revoked
Google issues manual actions specifically for structured data abuse, and losing eligibility is harder to reverse than earning it was. All four of these are avoidable.
Marking up content the page does not show
The rule is simple and strictly enforced: structured data must describe content visible on the page. FAQ markup for questions that appear nowhere, or prices the page does not display, is a violation regardless of whether the facts are true elsewhere.
Invented or aggregated review scores
The most common cause of a structured data penalty. Ratings must come from real reviews of the specific thing being marked up, displayed on the page. Self-serving markup where a business rates itself is explicitly disallowed.
Markup that contradicts the page
Hours in schema that differ from hours in the footer, or an address that differs from the contact page. This does not just fail to help — inconsistency reduces confidence in all your data, including the parts that were right.
Leaving it to rot after a change
Schema is written once and forgotten, then the business moves or changes its hours and the markup keeps asserting the old facts for years. Treat it as part of the content it describes, not as a one-time installation.
Common Questions
What people ask before implementing this.
Will adding schema improve my rankings?
Not directly — Google has said so repeatedly. It can increase clicks by making your result more prominent, and more clicks on a result that was already ranking is a real commercial gain. Expect a better result, not a higher one.
JSON-LD, Microdata or RDFa?
JSON-LD. Google recommends it, and it sits in a single block separate from your markup rather than being woven through your HTML, which makes it far easier to maintain and much harder to break during a redesign.
How do I check mine is right?
Google's Rich Results Test shows what your page is eligible for, and the Schema Markup Validator checks general validity. Test the live URL rather than a snippet, so that anything your template does differently in production is included.
Does schema help with AI search?
It helps with accuracy rather than being a requirement. Answer engines read the page content, but explicit machine-readable facts reduce the chance of a wrong detail — a mangled address or a misattributed quote — appearing in an answer that cites you.