← Back to guide info
4/23

Technical Foundations · 1.4

Schema Markup / Structured Data

Hidden code telling AI systems exactly what your business is, what it does, and where — the same schema markup covered in the SEO guide, but this time specifically for AI.

Why this matters

The customer’s side

Invisible directly, but the effect shows up as accuracy — a business with clean structured data is less likely to have an AI confidently state something wrong about it.

The AI’s side

Confirmed by Google, Microsoft, and ChatGPT directly

In March 2025, Google and Microsoft publicly confirmed they use structured data to help power their generative AI features, and ChatGPT has confirmed using the same kind of structured data too. This is one of the few GEO claims that's genuinely confirmed, not just observed or theorised.

See it happen

Same listing, same question asked. One of these tells the AI exactly what it's looking at.

No schema — plain result

rivingtons.co.uk

3-Bed Semi-Detached House For Sale, Oak Street

Guide price available on request. Contact us for viewing details and full particulars of this property…

Just a title and grey text. Nothing to catch the eye.

With schema — rich result

rivingtons.co.uk

3-Bed Semi-Detached House For Sale, Oak Street

★★★★★ 4.8 (127 reviews)

Guide price £450,000

3 bed · Semi-detached · Oak Street

Photo, rating, and price — pulled straight from the schema code.

How to check yours

Method 1 — Google's Rich Results Test

checks specifically for the schema types Google currently supports.

  1. 1.Go to search.google.com/test/rich-results.
  2. 2.Enter your page's URL and check the results.

Method 2 — Schema Markup Validator, for full validation

checks all schema.org markup for correctness, not just Google's supported rich-result types.

  1. 1.Open validator.schema.org.
  2. 2.Enter your page's URL and run the test.

How to fix what you find

No schema exists at all

generate it with AI and add it to your site's header — you don't need to write JSON-LD by hand.

  1. 1.Ask ChatGPT or Claude: "write LocalBusiness and Organization JSON-LD schema for [your business name], [address], [phone], [services], [opening hours]" — give it your real details.
  2. 2.Copy the code it gives you into the <head> section of your homepage, either directly or via your CMS's "custom code" or SEO plugin field.
  3. 3.Re-run the Rich Results Test to confirm it's now detected with no errors.

Schema exists but key details are missing or wrong

correct the specific fields the validator flags.

  1. 1.Note exactly which fields the Rich Results Test or Schema Markup Validator flags as missing or invalid.
  2. 2.Ask ChatGPT or Claude to add or correct just those fields in your existing code, pasting in what you currently have.
  3. 3.Replace the old code with the corrected version and re-validate.

You want to go beyond the basics

add FAQ and Service schema too, once Organization/LocalBusiness is solid — these give AI even more structured detail to draw on.

  1. 1.Ask ChatGPT or Claude to generate FAQPage schema from the actual FAQs already on your site, and Service schema listing what you offer.
  2. 2.Add both alongside your existing Organization/LocalBusiness markup, and validate all of it together.

Keeping on top of it

~15 min/month

Worth re-checking monthly, since site updates can silently break this without anything visibly changing.

How Reyse handles this

Done With You

we generate the code and tell you exactly where it goes.

Done For You

we implement and monitor this directly, ongoing.

Next up

llms.txt

Next →