Service Schema for AI Search: Organization, Author & Product Markup
Here’s a stat that seems to settle the schema debate once and for all: 65% of pages cited by Google AI Mode and 71% cited by ChatGPT include structured data (SE Ranking, 2025). Case closed, right? Schema equals AI visibility.
Not so fast.
A May 2026 study by Ahrefs, tracking 1,885 pages, tells a different story. When researchers added schema to pages already cited by AI systems, the results were underwhelming: AI Mode citations increased by just 2.4%, ChatGPT citations rose by 2.2%, and AI Overviews citations actually dropped by 4.6% (Ahrefs, 2026). The schema advocates and the schema skeptics are both missing the point.
Here’s what the data actually reveals: schema doesn’t boost citations for pages AI systems already see, but it’s essential infrastructure for being discovered in the first place. It’s the difference between wearing a name tag at a conference where everyone knows you, versus wearing one when you’re new in the room.
The schema landscape has also shifted dramatically. Google deprecated FAQ rich results in May 2026, following the earlier retirement of the HowTo schema. The schema types that drove rich snippets for years are now functionally decorative. What matters now are Organization, Author, Product, and Service schema, the structured data types that help AI systems understand who you are, what you offer, and why you should be trusted.
The stakes are significant, particularly for B2B companies. B2B Technology faces 70% AI Overview exposure, the highest of any industry (WebFX/Stackmatix, 2026). If you’re in professional services, enterprise software, or consulting, this isn’t optional anymore. AI Overviews now appear on 48% of all Google tracked queries, up from 31% in February 2025 (BrightEdge, 2026). The window for establishing entity authority is closing.
This guide covers what we use with clients at NAV43: the schema types that matter for AI search, how to implement them correctly with proper @id architecture, and how to avoid the implementation pitfalls that fragment your entity identity.
Why Your Schema Strategy Needs a Complete Overhaul
The schema playbook from 2023 is now obsolete. Google’s deprecation of FAQ rich results in May 2026 and the HowTo schema earlier that year marked the end of an era in which structured data primarily served as a rich result trigger. The new function of schema is fundamentally different: it’s how AI systems verify entity identity and establish trust.
Google’s March 2026 update made this explicit. AI Mode source selection now considers structured data quality alongside PageRank and content freshness. Schema has evolved from a ranking signal to an authentication layer.
The new schema priority stack looks like this:
- Organization schema establishes your entity foundation
- Author/Person schema verifies content creator credentials
- Service/Product schema defines your offerings
- Supporting schemas (BreadcrumbList, Speakable) enhance discoverability
This hierarchy matters because AI systems build understanding from the ground up. They need to know who you are before they can trust what you say, and they need to trust what you say before they’ll recommend what you sell.
The urgency is real. Organic CTR dropped 61%, from 1.76% to 0.61%, on queries where AI Overviews appear (Seer Interactive, 2025). The global GEO services market is projected to reach $1.48 billion in 2026 and grow past $17 billion by 2034 (DemandLocal, 2026). Companies are investing heavily because the search landscape has permanently shifted.
For a deeper dive into how AI systems evaluate sources beyond traditional ranking factors, see our guide on how answer engines choose sources.
The Contested Evidence: What the Data Actually Shows
Let’s address the elephant in the room: the evidence on schema’s impact is genuinely contested, and anyone telling you otherwise is oversimplifying.
The correlation evidence is strong:
– 65% of AI Mode cited pages include structured data (SE Ranking, 2025)
– 71% of ChatGPT cited pages include structured data (SE Ranking, 2025)
– Pages with structured data show 73% higher selection rates in AI Overviews (Wellows Research, 2026)
The causal evidence is weaker:
– Ahrefs’ controlled study found that adding schema produced minimal uplift for pages already being cited
– Only 38% of AI Overview cited pages rank in the top 10, down from 76% in mid-2025 (Ahrefs, 2026)
Here’s the critical nuance most content misses: the Ahrefs study only examined pages already heavily cited, those with 100+ existing citations. Schema may still be essential for pages that aren’t being discovered at all. Correlation studies showing high schema presence among cited pages suggest that schema is part of the baseline expectation for credible sources.
NAV43’s position: Schema is essential infrastructure, not a ranking lever. It helps AI systems understand what you are, not whether to cite you. For clients with common brand names or those operating in crowded verticals, we’ve observed schema helping with entity disambiguation, ensuring AI systems attribute content to the right organization.
What We’ve Seen at NAV43
One of our B2B technology clients shares a name with a consumer brand in an unrelated industry. Before implementing comprehensive Organization schema with proper sameAs links to their LinkedIn Company Page, Crunchbase profile, and industry directories, AI assistants regularly conflated the two companies. After implementation, brand attribution in AI responses improved significantly within 60 days. Schema didn’t boost their citation frequency, but it ensured citations went to the right entity.
The practical implication is clear: if your pages aren’t being cited at all, schema helps with discovery. If you’re already visible, schema helps with entity accuracy and trust.
Organization Schema: Your Entity Foundation
Organization schema is now the most critical structured data type for AI search. It’s how AI systems answer the fundamental question: “Who is this, and can I trust them?”
Essential properties for Organization schema:
| Property | Purpose | Example |
|---|---|---|
| name | Legal/official organization name | “NAV43 Inc.” |
| url | Primary website URL | “https://nav43.com” |
| logo | Official logo URL | “https://nav43.com/logo.png” |
| description | Concise organization description | “Canadian digital marketing agency…” |
| sameAs | Links to authoritative profiles | LinkedIn, Crunchbase, Wikidata |
| foundingDate | When established | “2014” |
| numberOfEmployees | Organizational scale | QuantitativeValue or range |
| address | Physical location | PostalAddress schema |
| knowsAbout | Topical authority areas | [“SEO”, “AI Search”, “MarTech”] |
The sameAs property deserves special attention. This is how you connect your Organization entity to the broader knowledge graph. AI systems use these connections for disambiguation, essentially verifying that you are who you claim to be. Link to:
- Your Wikidata entry (if you have one)
- LinkedIn Company Page
- Crunchbase profile
- Industry-specific directories
- Official social media profiles
The knowsAbout property is underutilized but increasingly important. This explicitly tells AI systems what topics your organization is qualified to discuss. It influences source selection when AI systems evaluate whether your content is authoritative on a given subject.
The results can be significant. InSinkErator saw a 69% increase in clicks for non-branded queries after implementing Entity Linking in their schema (Schema App, 2025). Entity linking through sameAs connections helped search systems, both traditional and AI, correctly identify and attribute their content.
For more on how entity-based optimization supports AI search visibility, explore our guide on entity SEO for AEO and GEO.
The @id Graph Architecture Most Marketers Miss
Here’s a technical detail that separates amateur schema implementation from professional-grade structured data: proper @id architecture.
The problem: Most sites fragment their Organization entity. The homepage has one version of Organization schema, the About page has a slightly different version, blog posts have yet another variation. To AI systems, these look like multiple different organizations.
The solution: Create a canonical Organization node with a stable @id that’s referenced site-wide. The @id acts as a unique identifier that tells AI systems, “All of these references point to the same entity.”
The standard format is: https://yoursite.com/#organization
This canonical Organization node should live on your homepage or a central template that appears site-wide. Every other page that references your organization, whether in Article schema, Service schema, or Author worksFor properties, should reference this @id rather than duplicating the full Organization schema.
Here’s how Author schema should reference your canonical Organization:
Notice that worksFor references the Organization @id rather than duplicating the full Organization schema. This creates a coherent entity graph rather than fragmented, potentially contradictory entities.
Practical implementation tip: Store your canonical Organization schema in a site-wide header component or template. Don’t copy-paste it per page, as this creates drift over time and inevitably leads to inconsistencies.
Author/Person Schema: E-E-A-T Signals for AI
AI systems don’t just evaluate organizational credibility; they evaluate individual author credentials. Author schema with proper expertise signals directly influences whether your content gets cited, particularly in YMYL (Your Money or Your Life) categories.
Critical properties for Author/Person schema:
| Property | Purpose | E-E-A-T Signal |
|---|---|---|
| name | Full author name | Trust |
| jobTitle | Professional role | Expertise |
| worksFor | Organization connection (@id) | Authoritativeness |
| knowsAbout | Topical expertise areas | Expertise |
| alumniOf | Educational credentials | Experience/Expertise |
| sameAs | LinkedIn, Twitter, profiles | Authoritativeness |
| award | Professional recognition | Expertise |
| hasCredential | Certifications/licenses | Expertise |
The knowsAbout property is particularly powerful for authors. This explicitly tells AI systems what topics the author is qualified to discuss. For a content strategist writing about SEO, knowAbout might include “Search Engine Optimization,” “Content Strategy,” “Keyword Research,” and “Technical SEO.”
Connecting Author to Article schema:
Every Article schema should reference its author using the @id pattern:
The author page strategy: Create dedicated author pages (e.g., /team/jane-smith/ or /about/jane-smith/) that serve as canonical sources for author entity data. These pages should include:
- Complete Author/Person schema
- Credentials and professional background
- Links to authored content
- External profile links (LinkedIn, industry profiles)
This approach builds a coherent author entity that AI systems can verify across multiple sources. For our complete framework on building machine-verifiable expertise, see our guide on author pages, E-E-A-T, and AI search visibility.
Complete Author schema example:
Service Schema: The B2B Opportunity Everyone’s Missing
Here’s a gap in the current schema landscape: most guidance focuses on Product schema for e-commerce or LocalBusiness schema for brick-and-mortar. Service schema for professional services, the bread and butter of B2B companies, is massively underutilized.
This matters because agentic AI browsers now compare options and conduct research on users’ behalf. Gartner projects 25% of organic search traffic will shift to AI chatbots and voice assistants by the end of 2026 (Gartner, 2026). When AI assistants evaluate service providers, Service schema provides the semantic foundation they need.
Service schema subtypes for professional services:
| Subtype | Use Case |
|---|---|
| ProfessionalService | Consulting, marketing agencies, advisory |
| LegalService | Law firms, legal consultants |
| FinancialService | Financial advisors, accountants, wealth management |
| EmploymentAgency | Recruiting, staffing services |
| MedicalBusiness | Healthcare providers, clinics |
| GovernmentService | Public sector services |
Critical properties for Service schema:
- name: Clear service name
- description: Detailed service description
- provider: Reference to Organization @id
- areaServed: Geographic coverage
- serviceType: Category of service
- offers: Pricing or engagement terms
- audience: Target customer segment
- serviceOutput: What the client receives
Service Schema Implementation for Professional Services
Let’s get specific with implementation examples for common B2B service types.
SEO/Marketing services example:
Service Schema Properties by B2B Service Type:
| Property | SEO/Marketing | Legal | Financial | Consulting |
|---|---|---|---|---|
| serviceType | “Search Engine Optimization” | “Corporate Law” | “Financial Advisory” | “Management Consulting” |
| serviceOutput | “Audit Report, Strategy” | “Legal Opinion, Contracts” | “Financial Plan, Analysis” | “Strategy Roadmap” |
| audience | B2B marketers | Corporations | HNW individuals, businesses | Enterprise executives |
| hasOfferCatalog | Tiered packages | Retainer options | Fee structures | Engagement models |
| areaServed | National/regional | Jurisdiction-specific | Regulatory region | Global/regional |
Multi-location considerations: If you offer services across different regions or markets, you have two options:
- Single Service with multiple areaServed values: For services delivered the same way everywhere
- Multiple Service entities: For region-specific service variations with different teams, pricing, or offerings
Both should reference the same canonical Organization @id to maintain entity coherence.
For guidance on building content that AI systems can quote and cite, see our GEO content strategy guide.
Product Schema: AI-Ready Product Markup
E-commerce sees AI Overviews on just 4% of queries, the lowest exposure of any industry (WebFX/Stackmatix, 2026). This might seem like Product schema is less urgent, but consider this: AI-driven retail search traffic grew 1,200%+ year-over-year (Adobe Analytics, 2025).
AI shopping assistants are growing rapidly, and they rely heavily on structured product data to make recommendations. Product schema remains essential for e-commerce, particularly for categories where AI assistants are becoming purchase advisors.
Critical Product schema properties:
| Property | Purpose | AI Impact |
|---|---|---|
| name | Product name | Direct citation text |
| description | Detailed description | Answer synthesis |
| brand | Brand reference | Entity linking |
| offers | Price, availability | Comparison data |
| aggregateRating | Review summary | Trust signal |
| review | Individual reviews | Social proof |
| sku | Stock keeping unit | Product identification |
| gtin | Global trade number | Universal identification |
Review and rating schema deserve special attention. AI systems place a heavy weight on social proof signals when making product recommendations. An aggregateRating with substantial review count signals market validation.
Note how the brand property references the Organization @id rather than duplicating brand information. This maintains entity coherence across your entire schema graph.
The NAV43 Schema Priority Framework
Here’s the tiered implementation approach we use with clients. It prioritizes the schema types that establish entity identity before moving to content-specific and supporting markup.
Tier 1: Foundation (Implement First)
These schemas establish who you are and must be in place before anything else:
- [ ] Organization schema with complete sameAs links (minimum 3-5)
- [ ] knowsAbout property with 5-10 specific topic areas
- [ ] Canonical @id established and documented
- [ ] Author/Person schema for all primary content creators
- [ ] Author worksFor properly referencing Organization @id
- [ ] Author knowsAbout aligned with content topics
Tier 2: Content Types (Implement Second)
These schemas describe what you offer and publish:
- [ ] Article schema for all blog/editorial content
- [ ] Article author property referencing Person @id
- [ ] Service schema for each major service offering
- [ ] Service provider referencing Organization @id
- [ ] Product schema (e-commerce) with offers and ratings
- [ ] Product brand referencing Organization @id
Tier 3: Supporting (Implement Third)
These schemas enhance discoverability and navigation:
- [ ] BreadcrumbList for site navigation
- [ ] Speakable for voice search optimization
- [ ] FAQ schema only for genuine FAQ pages (not for rich result gaming)
- [ ] HowTo schema only for actual instructional content
Validation steps:
- [ ] Test all pages with Google Rich Results Test
- [ ] Crawl site with Screaming Frog to audit schema consistency
- [ ] Verify @id references resolve correctly across pages
- [ ] Check for conflicting Organization entities
The logic behind this hierarchy: Foundation schemas establish entity identity that AI systems need to understand before they can trust you. Content-type schemas describe what you offer. Supporting schemas enhance discoverability but don’t define who you are.
For a comprehensive checklist on optimizing for AI search, see our complete AEO audit checklist for B2B websites.
Common Schema Pitfalls That Kill AI Visibility
After auditing hundreds of sites, we repeatedly see the same schema mistakes. Here’s what to watch for and how to fix it.
| Pitfall | Symptom | Fix |
|---|---|---|
| Fragmented entities | Different Organization schema on different pages with inconsistent data | Create canonical @id architecture; reference single Organization node site-wide |
| Missing sameAs links | Organization/Person schema with no external profile connections | Add minimum 3-5 sameAs URLs to authoritative profiles (LinkedIn, Crunchbase, Wikidata) |
| Orphaned Author schema | Author schema exists but doesn’t connect to Organization | Add worksFor property referencing Organization @id |
| Schema for deprecated rich results | FAQ/HowTo schema implemented primarily for rich snippets | Shift focus to Organization, Author, Service, Product; use FAQ only for actual FAQ pages |
| No knowsAbout properties | Missing topical authority signals | Add 5-10 specific topic entities relevant to your expertise areas |
| Over-marking content | Schema on every possible element, diluting signal quality | Focus on high-value pages and core entities; quality over quantity |
| Validation failures | Schema that doesn’t pass Google Rich Results Test | Test every page before deployment; fix errors immediately |
Fragmented entities is the most common and most damaging mistake. When AI systems encounter inconsistent Organization data across your site, they may treat these as different entities entirely, fragmenting your authority.
The fix: Document your canonical Organization @id and ensure every team member responsible for content or development knows to reference it rather than creating new Organization nodes.
Missing sameAs links is the easiest to fix and has an immediate impact. Without sameAs connections, AI systems have no external verification of who you are. This matters especially for organizations with common names or those in crowded verticals.
Orphaned Author schema breaks the chain of trust. An author might have excellent credentials, but if their schema doesn’t connect to a verified Organization, the credentials float without institutional backing.
For additional guidance on structuring content for AI systems, explore our guide on structured data for GEO.
Measuring Schema Impact on AI Visibility
Here’s the uncomfortable truth: there’s no “AI Overview impressions” metric in Google Search Console. Measuring a schema’s impact on AI visibility requires a different approach than traditional SEO measurement.
Manual auditing approach:
We run AI visibility audits for clients quarterly, tracking citation presence across 50+ priority queries. The process:
- Identify your top 30-50 target phrases based on search volume and business value
- Query each phrase in ChatGPT, Google AI Mode, and Perplexity
- Document whether your brand/content is cited, quoted, or referenced
- Track changes over time, correlating with schema and content updates
Proxy metrics to track:
- Branded search volume: Rising brand searches suggest increased AI-driven awareness
- Direct traffic: Users who heard about you via AI may visit directly
- “Mentioned in” brand monitoring: Tools like Mention or Brandwatch can track AI platform references
- Referral traffic from AI sources: Some AI platforms now pass referral data
Schema validation tools:
- Google Rich Results Test: Validates individual pages
- Schema.org validator: Checks schema syntax
- Screaming Frog: Site-wide schema audit capability
- Schema App: Enterprise-grade schema management
Set realistic expectations: Schema is infrastructure, not a quick-win tactic. Measure impact over quarters, not weeks. The goal isn’t immediate citation increases but building the entity foundation that enables long-term AI visibility.
For our complete framework on measuring AI search performance, see our guide on how to measure AI SEO.
Your Schema Action Plan Starts Now
Schema has evolved from a rich result trigger to an essential AI trust infrastructure. The types that drove visibility five years ago, FAQ and HowTo, are now deprecated. What matters now are Organization, Author, Service, and Product schema, the structured data types that help AI systems understand who you are and whether to trust you.
The data is nuanced. Schema doesn’t magically boost citations for pages AI already sees. But it’s essential for entity disambiguation, for discovery in the first place, and for ensuring that AI systems attribute content to the right organization.
Key takeaways:
- FAQ and HowTo schema are deprecated – Shift focus to Organization, Author, Service, and Product schema immediately
- Schema is infrastructure, not a ranking lever – It helps AI understand who you are, not whether to cite you
- The @id graph architecture is essential – Create canonical entity nodes referenced site-wide to avoid fragmenting your identity
- B2B companies need Service schema – It’s the most underutilized opportunity in professional services, and AI assistants are increasingly evaluating service providers
- sameAs and knowsAbout are your entity verification signals – Don’t skip them; they’re how AI systems confirm your identity and expertise
AI Overviews now appear on 48% of queries and are growing (BrightEdge, 2026). The window for establishing entity authority is closing. Companies that build coherent entity graphs now will have compounding advantages as AI search becomes the primary discovery channel.
Ready to audit your schema for AI search? NAV43’s GEO audit identifies exactly where your structured data gaps are and provides a roadmap for implementing fixes. Get your free growth plan and see where your entity foundation needs strengthening
