Schema Markup for AI Search: What's Changed in 2026
Schema used to be a Google ranking signal. Now it's how every AI engine reads your site. Here's what's still moving the needle in 2026.
The shift: schema is no longer optional.
Through 2023 you could rank fine without structured data, Google could parse your HTML and figure out what your page was about. Now four AI engines (Google AI Overview, ChatGPT, Perplexity, Bing Copilot) decide whether to cite you in seconds, and they lean heavily on schema to make that decision.
If you don't have schema, you're invisible to AI search regardless of how good your content is.
FAQPage: the highest-impact schema right now.
Of all the schemas, FAQPage earns the most AI citations relative to effort. AI engines parse it cleanly, and almost every page has Q&A content already (the "Common questions" section).
Wrap your existing FAQs in JSON-LD. Use complete sentences in the answer text. Aim for 4-6 questions per page. Don't repeat the same questions across pages, AI engines flag duplicate FAQ schema.
LocalBusiness with proper NAP, hours, and area-served.
For local businesses, LocalBusiness schema with full Name-Address-Phone consistency is the difference between being treated as a real entity and being treated as a placeholder. Add areaServed, openingHours, priceRange, and aggregateRating if you have it.
The availableLanguage array is underrated. If you serve Spanish or Korean speakers, list those languages. AI engines surface businesses to multilingual searchers based on this signal.
Article schema with explicit author.
Blog and editorial content benefits enormously from Article schema with an author field that's a real Person entity, not just a string. Link the author to an /about/ page that has the same Person entity.
This builds "authorship" that AI engines now weight heavily. A well-attributed article gets cited more than an anonymous one of the same quality.
Schemas that are table stakes (not differentiators).
BreadcrumbList, WebPage, WebSite, Organization: required for any modern site, but they don't earn citations on their own. Get them in place and stop optimizing them. Spend your structured-data effort on FAQPage, Article, and LocalBusiness instead.
New for 2026: ProfessionalService and Service-level pricing.
AI engines have started parsing Offer nested inside Service for price comparisons. If you list your pricing publicly (you should), wrap each package in Service with a nested Offer and a priceSpecification.
This is what makes "how much does a Bay Area marketing agency cost" queries cite specific businesses with specific prices, instead of citing generic articles.
Tracked queries, schema deployment, citation monitoring, monthly reporting. For Bay Area businesses serious about being cited in AI answers.
Talk to Eric →