The schema.org vocabulary lists over 800 types. Most websites need fewer than ten. The gap between what exists and what you should actually implement is where most of the confusion lives, so this guide covers the shortlist: the types worth knowing, where they apply, and which ones to prioritize first.
What schema markup does
Schema markup is structured data added to a page in a format search engines and AI systems can parse directly, without having to infer meaning from prose. The standard format is JSON-LD: a script block in the page’s <head> that declares what the page is, who published it, and how specific facts relate to each other.
The practical payoff runs in two channels. Certain types enable rich results in Google Search, things like star ratings beneath a product listing, an FAQ accordion under a how-to post, or a sitelinks search box beside your brand. The second channel matters as much now: structured data removes the ambiguity that costs a citation in AI-generated answers. Structured data for brands goes deep on that side of the value.
The realistic shortlist
Schema.org is intentionally exhaustive. It covers everything from MedicalCondition to VideoGame. Do not let the size of the vocabulary distract you. For most sites, the types that matter fit on one page.
Here is that shortlist, organized by where and why you would use each one.
Organization and WebSite
These two belong on every site, usually implemented in a global layout or the homepage. Organization tells engines who runs the site: name, logo, social profiles, and the relationships between entities you own. WebSite is what powers the sitelinks search box in Google when someone searches your brand directly.
Neither produces a rich result by itself, but both are foundational. If a model cannot confidently resolve who you are, it will not name you in a generated answer. Entity clarity starts here.
BreadcrumbList
BreadcrumbList marks up the navigation path above a page. It takes almost nothing to implement when your URLs already reflect a logical hierarchy, and it produces breadcrumb trails in search snippets. More importantly, it tells machines how your content is organized. It belongs on every page that sits below the homepage.
Article and BlogPosting
For editorial content, Article (or its more specific subtype BlogPosting) declares who wrote the piece, when it was published, when it was last updated, and what it covers. Search engines use the modified date when judging freshness. AI answer engines treat author, publisher, and publication date as trust signals when deciding whether a passage is worth quoting.
Include datePublished, dateModified, author, and publisher at minimum. If you have an author bio page, link to it with a Person entity under author. That makes authorship machine-verifiable, which carries weight as content credibility becomes more scrutinized.
FAQPage
FAQPage is one of the highest-return schema types for informational content. It marks up a list of questions and answers on a page, and Google has used it to render expandable FAQ blocks in search results. Even when the rich result does not appear, AI answer engines can extract clean, self-contained question-answer pairs from it without guessing at the structure.
If your page already has a FAQ section, not adding
FAQPageschema leaves structured, quotable content invisible to machines.
Worth noting the tradeoff: Google has pulled back on displaying FAQ rich results for some content categories. The schema is still worth adding for AI citation purposes, but do not build a page around it expecting guaranteed SERP expansion.
Product and AggregateRating
For e-commerce or SaaS pages, Product is the type that earns star ratings, price, and availability in search results. Pair it with AggregateRating when you have genuine review data. Do not inflate ratings or manufacture reviews; Google’s guidelines here are strict and penalties are visible.
Offer nests inside Product to describe pricing and availability. If you sell software with a free tier, SoftwareApplication is a closer fit than bare Product.
LocalBusiness
If your business has a physical presence, LocalBusiness (or a subtype such as Restaurant, MedicalBusiness, or LegalService) is mandatory. It feeds the knowledge panel in Google and contributes to maps listings. Include name, address, phone, hours, and geo coordinates. Consistency between this schema and your Google Business Profile matters; mismatches introduce doubt.
HowTo and Event
HowTo marks up step-by-step instructional content with discrete steps. Add it when a page genuinely walks a reader through a process. Event applies to any page about a specific event with a start date, location, and organizer. If the event moved online, set eventAttendanceMode to OnlineEventAttendanceMode.
Both types can produce rich results in search, though display frequency varies by vertical and changes over time.
Choosing which types to add to each page
The decision is simpler than the type count suggests. Start from your page types:
- Every page:
OrganizationandWebSite(global),BreadcrumbList - Blog and editorial:
ArticleorBlogPosting,FAQPagewhere the content supports it - Product or pricing pages:
Product,Offer,AggregateRating(with real review data) - Location pages:
LocalBusiness - Instructional content:
HowTo - Event listings:
Event
A common mistake is adding every possible type to every page. That introduces noise. Be specific: add the type that accurately describes the page. A blog post is not a Product. A pricing page is not an Article.
One more thing to get right: always implement schema as JSON-LD rather than Microdata or RDFa. JSON-LD lives in the <head>, separate from your HTML, and is easier to maintain and audit. Use the Rich Results Test to validate; schema errors are quiet and do not break pages, they just silently fail to produce rich results.
The technical SEO checklist covers schema implementation alongside the other foundational items worth reviewing together, and the on-page SEO guide places schema in the context of the full on-page signal set.
How Strynal approaches schema markup types
We implement schema as part of a broader AI visibility practice rather than a one-off task. The types you add, how you nest entities, and whether your Organization connects correctly to your Article authors all compound. Schema that contradicts other page signals produces confusion rather than clarity.
For each site, we map the page types first, write JSON-LD that fits the actual content, validate it, and wire it into the CMS so editors do not have to think about it. The goal is structured data that stays accurate as content changes, which is the only kind that helps over time.
If your site’s schema has never been audited, or your team added it once and moved on, it is worth checking. Stale or incorrect schema can send the wrong signals at the exact moment a search engine or AI model is deciding whether to trust you.