What is schema markup?

Schema markup is structured data added to web pages that tells search engines and AI engines what the content represents. Implemented as JSON-LD in the page head, it makes content machine-readable in ways the page text alone isn't.

Which schemas should we implement?

For most operator-run businesses: Organization on every page, Service on each service page, Article on each blog post, FAQPage where Q&A content exists, HowTo for procedural content, BreadcrumbList for nested pages, and Person for author bylines. These cover the majority of high-value cases.

Does schema markup actually help SEO?

Yes, in two ways. It enables rich results in classic search (FAQ accordions, HowTo cards, etc.) and it provides explicit signals that AI answer engines use when deciding what to cite. The impact is larger for AEO than for classic SEO.

Schema markup: the part of SEO that actually still matters

Schema markup is the part of SEO that became more valuable when answer engines arrived. For classic SEO, schema enabled some rich results — useful, not decisive. For AEO, schema is structurally important: it tells the AI what the content represents in a machine-readable way that page text alone can’t.

This is what schema actually is and which schemas are worth implementing for an operator-run business.

What schema markup actually is

Schema markup is structured data added to a web page in a format the engines can parse directly. The most common implementation is JSON-LD — a block of JSON in the page’s <head> that describes what the page represents.

A simple example:

{
  "@context": "https://schema.org",
  "@type": "Organization",
  "name": "Example Business",
  "url": "https://example.com",
  "description": "What the business does."
}

That block tells any consumer of the page (search engine, AI engine, browser, future tool not yet invented) that the page represents an organization with this name and description. The page text might say the same thing, but the schema makes it unambiguous and machine-readable.

The vocabulary comes from schema.org, a shared standard supported by Google, Bing, Yahoo, Yandex, and increasingly the major AI engines. Implementing schema is implementing the shared standard.

Why schema matters more for AEO than for SEO

For classic search, schema enabled rich results — the visual enhancements that appear in search results pages. FAQ accordions, HowTo cards, review stars, recipe details. Useful but optional; sites without schema still ranked.

For AEO, schema is more structurally important. AI answer engines rely on:

Knowing what entities the content describes (Organization? Person? Service? Article?)
Knowing how content relates (this article was written by this person, who works for this organization)
Knowing what specific data points the content contains (this is a price; this is a date; this is a step)

Without schema, the AI has to infer this from the text, and inference is conservative. With schema, the AI has explicit signals it can rely on. Sites with strong schema implementation get cited more often than equivalently-good sites without.

The schemas that matter most

For an operator-run business, seven schemas cover the majority of high-value cases.

Organization

The foundational schema. Every page should reference an Organization. Include:

Name, URL, logo
Description
sameAs array linking to social profiles, LinkedIn company page, etc.
contactPoint with how to reach the business
knowsAbout listing the expertise areas the business covers

The Organization schema is the entity anchor — other schemas reference it via the Organization’s @id. Get this one right; it pays back across everything else.

Service

For each major service offering, a Service schema with:

Name, description, serviceType
provider linked to the Organization (@id)
audience describing who the service is for
areaServed if relevant

Service schema is what tells the AI which services the business offers and to whom. This is one of the highest-leverage schemas for vendor-recommendation queries.

Article (or BlogPosting)

For each blog post:

headline, description, image
datePublished and dateModified
author (as Person or Organization)
publisher linked to the Organization
mainEntityOfPage linking back to the canonical URL

Article schema makes blog content discoverable and citable. AI engines pay particular attention to date fields when assessing recency.

FAQPage

For pages with Q&A content:

An array of Question items
Each with name (the question) and acceptedAnswer (the answer)

FAQ schema enables rich results in classic search and provides clean extraction targets for AI engines. Use it on service pages, pricing pages, and blog posts with explicit Q&A blocks.

HowTo

For procedural content (numbered steps, recipes, instructions):

Name and description of the procedure
An array of HowToStep items
Each step with name and text

HowTo schema enables rich step-by-step results in classic search and gives AI engines explicit step structure they can cite individually.

BreadcrumbList

For nested pages (/services/websites/pricing etc.):

An array of ListItem items showing the path
Each with position, name, and URL

BreadcrumbList helps both search and AI engines understand site hierarchy and navigate the content structure.

Person

For author bylines:

Name, optional jobTitle, optional affiliation
sameAs linking to LinkedIn, professional bio, etc.

Person schema connects content to humans, which builds authority for AI engines and supports the broader entity graph.

What to skip (for most operators)

Schema.org has hundreds of types. For most operator-run businesses, several are less useful than they look:

AggregateRating without real reviews. Faking ratings via schema is detectable and counterproductive.
LocalBusiness when the business isn’t local-first. Use Organization for geographically-neutral businesses.
Schema for pages that don’t have the underlying content. Schema describes content; describing content that doesn’t exist confuses the engines.
Highly specialized schemas without need. SoftwareApplication, Recipe, Event, etc. are valuable in their domains but irrelevant for most service businesses.

The principle: implement schemas that match what your pages actually represent. Don’t implement schemas because they exist.

Common implementation mistakes

Five patterns that produce broken schema:

Schema without supporting page content. FAQ schema on a page with no actual FAQ content. The engines detect this and devalue the source.

Inconsistent entity references. The Organization name in Organization schema doesn’t match the name in Service schema. The engines can’t tell whether they’re the same entity.

Stale dates. Article schema with datePublished and dateModified that aren’t actually accurate. The engines use these as signals; wrong dates produce wrong signals.

JSON-LD syntax errors. A misplaced comma or quote breaks the entire schema block. Always validate before deploying.

Multiple conflicting schemas. Two FAQ blocks on the same page with overlapping questions. Pick one canonical instance per page.

Validating schema before it ships

Three tools for validation:

Google’s Rich Results Test — checks whether the schema is recognized by Google specifically and whether it’s eligible for rich results
Schema.org’s validator — checks against the schema.org vocabulary itself
Manual inspection of the JSON-LD — confirms that the structure is what you intended

Run all three before deploying schema changes. Errors prevent recognition; warnings often signal real issues.

What “we handle” looks like at this layer

For operators implementing schema across an existing site:

Audit of current schema coverage and quality
Implementation of the seven core schemas across all relevant page types
Validation against the major engines’ tools
Consistent entity referencing across all schemas (one Organization, referenced from everywhere)
Schema updated as content changes (dates, prices, services)
Schema layer maintained as standards evolve

This isn’t a one-time project — schema standards evolve, the engines update what they look for, and content changes mean schemas need updates. Implementation plus tending is the right shape.

A practical first step

If you want to assess where you stand:

Open your homepage. View source. Search for application/ld+json.
If you find schema, copy it into Google’s Rich Results Test. See what it says.
If you don’t find any schema, that’s the gap. Schema is one of the highest-leverage interventions you can make on a site that’s currently missing it.

For most operator-run sites we audit, schema is either absent, partial, or implemented inconsistently. The gap between “no schema” and “complete, validated schema” is one of the cleanest improvements in modern SEO/AEO work — straightforward to implement, hard to manipulate, and meaningful in impact.