Technical Signals · P1.2
Structured data for AI search: the four schemas that actually move the needle
tl;dr
Schema markup tells AI assistants what each page is about in a format they don't have to guess at. Four schemas cover 90% of the value: Organization sitewide, plus Article, FAQPage, or Product per page type.
// your move
See what AI assistants say about your site.
Free, 30 seconds, no signup. We'll check all 15 signals and tell you which ones to fix first.
Run the auditWhat schema actually does
Schema is structured data written in JSON-LD and dropped into a script tag in your HTML. It's a machine-readable summary of the page. AI assistants love it because it removes guesswork.
Without schema, an assistant has to infer that your /about page describes your company. With Organization schema, you just hand it the founding year, address, logo, and social profiles.
Schema #1: Organization (every page, sitewide)
Goes in your site header or footer. Identifies who you are, where you're based, and how to reach you. Reuse the same block on every page.
- 01Set @type to Organization (or LocalBusiness if you have a physical location).
- 02Fill in name, url, logo, sameAs (array of your social profile URLs), and contactPoint.
- 03For local businesses, add address, geo (latitude/longitude), and openingHours.
- 04Validate at validator.schema.org before publishing.
Schema #2: Article (every blog post, news piece, guide)
Wraps each article with author, publish date, modified date, and headline. AI assistants use the date to decide whether your content is fresh enough to cite.
- 01Set @type to Article (or NewsArticle for news, BlogPosting for blog posts).
- 02Include headline, datePublished, dateModified, author (with @type: Person and name), and image.
- 03Add description matching your meta description.
- 04Keep dateModified honest. Lying about freshness gets noticed.
Schema #3: FAQPage (any page with Q&A)
If your page has questions and answers — and it should — wrap them in FAQPage schema. This is the single highest-ROI schema for AI assistants because it maps cleanly to how they answer questions.
“Wraps each article with author, publish date, modified date, and headline.”
// tool we'd actually reach for
Schema Pro
Add schema markup to your WordPress site.
Check out Schema Proaffiliate link — we may earn a commission
Schema #4: Product (e-commerce)
For each product page: name, image, description, brand, offers (price, currency, availability), and aggregateRating. Assistants pull these into shopping answers and comparisons.
How to add it without breaking things
- 01Pick one page type (start with Article).
- 02Write the JSON-LD block. Use a generator like merkle.com/schema-markup-generator if you're new.
- 03Paste the <script type="application/ld+json">...</script> into the page head or just before </body>.
- 04Validate at validator.schema.org. Fix any errors.
- 05Test in Google's Rich Results Test for SEO bonus points.
- 06Repeat for the next page type.
Common questions
Do AI assistants actually parse JSON-LD?
Yes. It's the same format Google has used for a decade. ChatGPT, Perplexity, and Claude all extract it when they fetch a page.
Can I just use Yoast or Rank Math?
Yes, and you should if you're on WordPress. They handle Organization and Article schema automatically. You can add FAQPage and Product manually or via a paid tier.
What if I have duplicate schema?
Multiple JSON-LD blocks on one page are fine if they describe different entities. Don't define the same Organization twice on the same page.
Keep reading
// your move
See what AI assistants say about your site.
Free, 30 seconds, no signup. We'll check all 15 signals and tell you which ones to fix first.
Run the audit ⚡