Schema Markup for AI Search: What Helps and What Is Hype
TL;DR: Schema markup is structured data that labels what a page is about. It helps search engines understand your content and can unlock rich results, but Google says AI Overviews and AI Mode need no special schema, and no markup guarantees an AI citation. Use accurate JSON-LD for Organisation, LocalBusiness, Article and Service, keep it matching the visible page, and validate it. Treat it as clarity, not a trick.
Does schema markup help AI search?
It helps machines read your page more accurately, but it does not guarantee that an AI assistant will cite you. Google explicitly says no special schema is required for AI Overviews or AI Mode.
Google’s page on AI features in Search states that there is no special schema.org structured data you need to add to appear. It asks that any structured data you do use matches the visible text of the page.
Structured data still has a clear job. Google describes it as giving explicit clues about the meaning of a page, and it can make pages eligible for rich results.
For other AI products, the picture is less documented. A model that reads page text may use markup as a cleaner signal of who you are and what you offer, but this is a reasonable inference, not a published rule. Plan for clarity and accept uncertainty.
Key takeaway: schema is a clarity and eligibility tool. It supports good content and does not replace it.
Which schema types matter for a UK business?
Start with Organization, LocalBusiness (where you serve a local area), the page-level types that fit your content such as Article and Service, and BreadcrumbList for site structure.
The vocabulary comes from schema.org, whose types use US spelling, so write Organization in code even though British English is used on the page.
| Type | Use it for | Where |
|---|---|---|
Organization | Name, logo, URL, contact details, social profiles | Home and about pages |
LocalBusiness | Address, opening hours, service area | Home, contact and location pages |
Service | A specific service, provider and area served | Service pages |
Article | Headline, author, dates and publisher | Guides and blog posts |
BreadcrumbList | Page hierarchy | Inner pages |
Person | Author or team member, linked to credentials | Author and about pages |
The strongest use is entity clarity. If your Organization markup, your Google Business Profile and your page text all give the same business name, address and phone number, systems have fewer conflicts to resolve.
Use the sameAs property to link to profiles that genuinely belong to the business, such as Companies House, LinkedIn or Google Business Profile. Only include profiles you control or that officially describe you.
What format should you use?
Use JSON-LD. Google recommends it where your site setup allows, and it keeps markup separate from the visible HTML.
Google also accepts Microdata and RDFa if they are valid. JSON-LD is easier to write, audit and maintain because it sits in one script block rather than being spread across the page.
On WordPress, SEO plugins such as Rank Math generate much of this automatically. Check the output rather than trusting it, because defaults often describe the wrong organisation type or leave out the fields that matter.
Keep the markup honest. Every fact in it should be visible to a person reading the page, whether it is a price, an author, a date or an address.
What happened to FAQ schema?
Google now restricts FAQ rich results, so FAQ markup no longer produces the visible search enhancement for most sites. The FAQ content itself still matters, because clear question-and-answer passages are easy to quote.
Check Google’s current FAQPage documentation for the eligibility rules before adding it for rich results. Do not rely on memory, as search features change.
The value of an FAQ section is in the writing. Short, self-contained answers that name their subject are cheap for any system to extract, with or without markup.
Adding FAQPage markup where the page genuinely contains those questions is harmless if valid. Expecting it to create a special route into AI answers is not supported by anything Google has published.
What are the common schema mistakes?
The most common mistakes are marking up content that is not on the page, copying another business’s markup, and leaving contradictory data across plugins.
| Mistake | Why it matters | Fix |
|---|---|---|
| Markup differs from visible text | Breaks Google’s guidance and erodes trust | Generate markup from the same source as the page copy |
| Fake or inflated review markup | Risks a manual action and misleads users | Mark up only genuine, visible reviews |
| Two plugins output duplicate Organization data | Conflicting entities | Disable one source of schema |
| Wrong business type | Misdescribes what you do | Choose the most specific accurate type |
| Stale details after a move or rebrand | Old address or name persists | Review markup in every site change |
Another frequent problem is treating schema as the whole job. A page with perfect markup but a vague first paragraph and no clear answer still gives a system little to quote.
How do you validate and maintain it?
Test each template with Google’s Rich Results Test, check the schema.org validator for syntax, and monitor Search Console’s rich result reports after launch.
Test before launch
Paste the URL or code into the Rich Results Test and fix any errors or warnings that apply to your type.
Inspect after launch
Use URL Inspection in Search Console to confirm Google found the markup on the live page.
Monitor
Watch rich result status reports, because template changes can quietly break markup.
Review quarterly
Check addresses, opening hours, names and author details against the live site.
If you want help building entity-consistent markup as part of a wider visibility plan, our answer engine optimisation services cover structured data alongside content and technical checks.
For the surrounding strategy, read entity SEO for UK businesses, how AI Overviews choose sources and the llms.txt guide.
Key takeaway: accurate, consistent JSON-LD is worth doing. Promising AI citations because of it is not.
Schema markup FAQs
Does schema markup get you cited in AI Overviews?
There is no evidence that it does. Google says AI Overviews and AI Mode need no special schema. Pages must be indexed and eligible to show with a snippet. Schema can still help Google understand the page.
Which schema format does Google recommend?
Google generally recommends JSON-LD where your setup allows it. Microdata and RDFa are also accepted when valid. JSON-LD is usually simplest to maintain because it lives in one block rather than throughout the HTML.
Is schema.org markup a ranking factor?
Google does not describe structured data as a ranking factor. It describes it as a way to help Google understand a page and to qualify for rich results. Do not expect markup alone to move rankings.
Should a small UK business use LocalBusiness schema?
Yes, if it serves customers in a defined area or has a premises. Include accurate name, address, phone number and opening hours that match your Google Business Profile and visible page. Consistency matters more than volume of properties.
How do I check my structured data is valid?
Use Google’s Rich Results Test and the schema.org validator, then URL Inspection in Search Console for live pages. Check rich result reports regularly, since theme or plugin updates can break markup without warning.
Can schema markup hurt my site?
Yes, if it is misleading. Markup that does not match visible content, or fake review data, can lead to lost rich results or a manual action. Keep it accurate and generated from the same source as the page copy.
Do AI assistants other than Google use schema?
It is not documented for most of them. Schema may help any system that reads the page, but vendors have not confirmed it as a citation signal. Prioritise clear text, crawl access and consistent facts, then add schema.