Creative Works Schema

TL;DR — “Creative Works schema” is my shorthand for the schema.org types you use to mark up content you’ve created — articles, videos, podcasts, recipes, courses, datasets, movies, and books. There are eight in this cluster. But here’s the catch most guides hide: adding one of these does not guarantee a fancy search listing. Some specific subtypes support documented Google features, while others do not. The range runs from mainstream visual treatments to narrow features or nothing you can see at all — the markup just helps engines understand your page. Valid is not the same as visible.

What “Creative Works schema” means

When you look at a recipe page, you can tell the cook time from the calories from the star ratings just by reading. A search engine sees plain text and has to guess. Schema markup is code you add to spell it out — “this is the cook time,” “this is a video,” “this is the author” — using a shared vocabulary from a site called schema.org.

“Creative Works schema” is my umbrella for the schema.org types that describe a piece of authored content. schema.org really does group these under one formal type called CreativeWork — but Google does not treat them as one family, and I’ll be honest about that in the Advanced tab. The eight I cover here:

• Article — a news story, blog post, or article.
• AudioObject — an audio file, like a podcast episode.
• Book — a book.
• Course — an educational course.
• Dataset — a downloadable data set.
• Movie — a film.
• Recipe — a cooking recipe.
• VideoObject — a video.

The thing most people get wrong

Here’s the myth this whole hub exists to bust: not every schema type gives you a visible rich result. People assume “add schema → get a fancy listing.” That’s only true for a couple of these.

• Recipe and Video — yes, these earn the enhanced listings you picture (stars, cook time, video thumbnails). Prioritize these if you publish that kind of content.
• Course, Movie, Dataset — real features exist, but they’re narrow: a course carousel (needing at least three courses), a mobile-only movie carousel, and datasets that show up in a separate Google tool, not normal search.
• Article, AudioObject, Book — no dedicated visual card in regular search. The markup helps search engines understand the page more precisely, but you won’t see a special listing from it. I don’t have verified evidence of a separate AI-specific benefit beyond that, so I’m not going to claim one.

So valid markup is worth doing for understanding — but don’t expect every type to change how your listing looks.

One more thing it does not do

Schema markup does not make you rank higher. Google has said this plainly and repeatedly. What it does is make you eligible for a richer listing (where one exists) and help engines understand your content. That’s different from ranking better.

Want the full version — the tier-by-tier accuracy table, what changed in June 2025, and which of the eight applies to your content? Switch to the Advanced tab.

TL;DR — “Creative Works schema” is my practitioner grouping for eight of the schema.org CreativeWork subtypes that mark up authored content — Article, AudioObject, Book, Course, Dataset, Movie, Recipe, VideoObject. CreativeWork is a real schema.org umbrella type (Thing > CreativeWork) with dozens more subtypes than these eight, and Google does directly recognize the base CreativeWork type in one narrow context — its paywalled-content structured-data profile — so it’s wrong to say Google never touches the type. What Google doesn’t do is group these eight into one feature family: each gets its own standalone guide or none. The differentiating fact is feature-specific support, not membership in the CreativeWork hierarchy. Course is now a 3+-course carousel only (the single-course card was deprecated June 2025); Movie is a mobile-only carousel (and its Critic Review layer was deprecated); Dataset feeds the separate Dataset Search tool, confirmed by Google in a November 2025 clarification; Article has no dedicated visual card and does not require AMP; AudioObject and Book have no supported-types entry for a mainstream rich result (verified against Google’s current gallery, 2026-07-16). Valid ≠ visible. And none of it is a ranking factor — it buys eligibility, not rank.

First, an honest framing: this is my grouping, not Google’s

I want to be transparent up front, because most content that talks about “creative works” or “content” schema quietly implies these types are an official Google family. They’re not — with one narrow exception worth naming precisely.

• schema.org really does group them, and my eight aren’t exhaustive. CreativeWork is a genuine, formal schema.org type, and its hierarchy (Thing > CreativeWork) files dozens of subtypes under it — Article, Book, Course, Dataset, Movie, MusicRecording, Painting, Photograph, Play, Recipe, ShortStory, SoftwareApplication, Thesis, VideoObject, AudioObject, and more. This hub covers the eight that matter most for on-page SEO practice, not the full list, so treat it as a curated slice rather than a complete map of the type.
• Google doesn’t group them into a feature family — but it does use the base type in one bounded case. Google’s per-type rich-result documentation never mentions a “creative works” umbrella; each type gets its own standalone guide (or no guide at all), with zero cross-linking between them. But Google’s paywalled-content structured data guidance directly lists CreativeWork (the base type) and several subtypes as supported page types for that narrow profile. So “Google never uses CreativeWork” is false; the accurate statement is that Google doesn’t treat these eight as one rich-result family. And schema.org’s own CreativeWork page is pure vocabulary — it never mentions Google, SEO, or rich results at all.

So why put these eight in one article? Because in practice they’re the same kind of problem: “I publish X kind of content — which schema type applies, and will it actually get me a visible rich result?” That question is real and useful. The claim that Google rewards “creative works schema” as one feature category is not. I’d rather say that plainly than pretend Google drew this box.

The tiered accuracy table: which of the eight actually get a visible result

This is the single thing this hub does that almost no competing source gets right. “Valid schema” and “visible rich result” are not the same, and for these eight types the gap is enormous — it ranges from a mainstream SERP card all the way down to zero visual treatment. Below is a dated consumer/profile matrix rather than an opaque tier label — each row names the actual Google consumer, its gate, and the date I last confirmed it against Google’s current Search Central docs (all rows verified 2026-07-16). One rule holds across every row: the absence of a rich-result feature for a type is not the same as a proven benefit elsewhere (I’m not going to convert “no visual card” into an unmeasured AI-understanding claim just because it sounds better).

The one sentence to internalize: valid schema for a CreativeWork subtype does not imply a visual rich result — it ranges from a mainstream SERP card (Recipe, VideoObject) down to zero visual treatment at all (AudioObject, Book). Conflating “valid” with “visible” is the most common mistake in this space, and it’s the flaw that runs through almost every competing guide.

What changed recently: June 2025 was an inflection point

If you’re working from older advice, three things in this cluster changed and they matter. Each item below has its own feature name and date so you don’t conflate a retired feature with the related one that survives — verified against Google’s Search Central update log, 2026-07-16:

1. Course Info card — announced for deprecation June 12, 2025 (removed). The richer single-course card (price, dates, instructor) was one of seven types removed in Google’s structured-data simplification, on the reasoning that these features “weren’t commonly used in Search.” This is a different, separately named feature from Course list, which remains documented and current — it requires 3+ courses marked up together with carousel markup. Any guide promising you a single-course rich card is now stale. (I confirmed the old course-info doc URL now redirects to Google’s changelog announcing the removal.)
2. Critic Review layer — deprecated. The separate Critic Review markup (used by Movie, Book, and local business) was retired. Google’s framing, relayed through industry coverage, was that it “wasn’t useful for the ecosystem at scale,” and it noted the markup is harmless to leave in place for understanding. This is a different feature from the Movie carousel, which survives, mobile-only, and remains separately documented.
3. Book Actions — announced for deprecation June 2025, deprecation banner removed November 5, 2025. Book Actions (a registration-gated, feed-based “buy/borrow” program — not simple on-page JSON-LD) was in the June 2025 announcement batch, was excluded from the later removal wave, and Google subsequently took the deprecation banner off the live Book doc on November 5, 2025. So don’t state Book Actions was removed — it’s a narrow, gated, still-live feature. (More in the Book deep dive.)

The Book and AudioObject nuance — the two “no visible result” cases

These two are where competitors are vaguest, so I want to be precise:

• AudioObject — there’s no dedicated rich-result guide in Google’s current supported-types gallery (verified 2026-07-16). Google’s podcast and media-clip structured-data doc URLs both 404, and AudioObject doesn’t appear on Google’s supported-structured-data-types list (the “search gallery”). Google’s separate podcast guidance is about RSS-feed podcast management, not schema.org AudioObject markup. That absence is limited to Google’s gallery, though — it isn’t proof that no system anywhere consumes the markup, and I don’t have verified evidence of a specific alternative benefit (AI or otherwise) to claim in its place. Mark it up for vocabulary correctness and entity understanding, not for an expected search feature.
• Book — plain Book schema has no dedicated mainstream rich-result card either. The nuance is Book Actions: a narrow, registration-gated feed program that was announced for deprecation in June 2025 but was not actually removed, and remains live (deprecation banner removed November 5, 2025). Book is also listed as a reviewable entity type, so it can feed Google’s general review-snippet feature. But “add Book schema and get a search feature” is not true for the on-page JSON-LD most people mean.

The Article correction: no AMP required, no dedicated card

Article (which covers Article, NewsArticle, and BlogPosting in one Google guide) is the type with the most stale advice attached to it, so two corrections:

• No dedicated visual card. Google frames the benefit as better title/image/date display and general eligibility (Google News, Assistant), not a distinct snippet type. There are also no required properties — you add the properties that apply.
• Article schema does not require AMP. There is no AMP requirement anywhere in Google’s current Article guide. That’s legacy advice worth explicitly correcting when you see it — and you’ll see it a lot.

How to decide which type(s) apply to your content

Do this in order: first pick the most accurate CreativeWork subtype for what the content actually is — don’t reach for a broader or better-known type because it sounds more likely to earn a feature. Only after the subtype is settled should you check which consumer-specific eligibility applies to it. Two questions do most of that second step: (1) what kind of content is this, and (2) do you actually want a visible result or just vocabulary/entity understanding? The Decision Trees tab walks it step by step, but the shape is:

• Recipe / Video content → mark it up; these earn real visual results. Highest ROI.
• A set of 3+ courses / a movie / a dataset → mark it up if the narrow feature fits (course carousel, mobile movie carousel, Dataset Search), knowing the reach is limited.
• Articles, podcasts, books → mark them up for entity understanding and general eligibility — not a verified AI-specific benefit — but don’t expect a special listing to appear.

Shared ground rules across all eight

Even though Google doesn’t treat these as one family, they share Google’s general structured-data guidelines, and the rules are worth stating once:

• Required properties are the eligibility gate — where a rich result exists at all. Miss a required property and you’re not eligible for that feature.
• Eligibility ≠ guaranteed display. Valid markup gets you into the pool; Google’s systems still separately decide whether to show the enhancement.
• Only mark up visible, accurate content — no invisible or misleading markup.
• JSON-LD is the recommended format for all of them — easier to maintain at scale than inline Microdata/RDFa.
• It’s not a ranking factor. Structured data earns eligibility for a feature and aids understanding — it never lifts your position.

Where to go next

This hub is the map; each of these is its own deep dive nested under it. I’ve written the one-line pitch to reflect each type’s actual tier — I’m not going to oversell the weak ones:

• Article schema — Article/NewsArticle/BlogPosting in one guide. No required properties, no dedicated visual card, and (importantly) no AMP requirement — read it as much for the myth-correction as the implementation.
• AudioObject schema — podcast and audio markup for vocabulary and entity understanding, with no verified AI-specific benefit claimed; there’s no mainstream rich result here, and the deep dive is honest about that.
• Book schema — no dedicated mainstream card, plus the full, careful story of Book Actions (announced-for-deprecation but not removed) so you don’t repeat the common error that it’s gone.
• Course schema — the current reality: a 3+-course carousel, not the single-course card that was deprecated in June 2025.
• Dataset schema — feeds Google’s separate Dataset Search tool, not the normal SERP; worth it if data discovery is your goal, not a classic snippet.
• Movie schema — a mobile-only carousel, and why the separate Critic Review layer is gone. Set expectations before you invest.
• Recipe schema — one of only two types here with a mature, mainstream visual result (stars, cook time, calories). High priority if you publish recipes.
• VideoObject schema — the other Tier 1 type: thumbnails, duration, and key moments across the main SERP, Video tab, Images, and Discover. High priority for video.
• ImageObject schema — rarely a destination on its own; its main job is as the nested image/logo value inside Product, Article, Recipe, and Organization, plus the one visible feature it does unlock directly: Google’s Licensable badge.

For the wider picture, this hub nests under the broader structured data subcluster — see also the sibling schema markup article there for the vocabulary, formats, and deprecation cycle. And it all lives in the on-page cluster, where structured data sits alongside the rest of on-page technical SEO.

AI summary

A condensed take on the Advanced version:

• What it is: “Creative Works schema” is Patrick’s practitioner grouping for eight of the schema.org CreativeWork subtypes that mark up authored content: Article, AudioObject, Book, Course, Dataset, Movie, Recipe, VideoObject (the schema.org hierarchy has more; these eight aren’t exhaustive).
• Half-official: CreativeWork is a real schema.org umbrella type (Thing > CreativeWork). Google does directly recognize the base type in one narrow context — its paywalled-content structured-data profile — so it’s inaccurate to say Google never uses it. What Google doesn’t do is group these eight into one rich-result category: each type gets its own standalone guide or none, and schema.org’s page never mentions Google/SEO/rich results.
• The core value — valid ≠ visible. Dated consumer/profile matrix (verified 2026-07-16):
  • Full mainstream visual result: Recipe (stars/cook time/calories), VideoObject (thumbnail/duration/key moments; SERP, Video tab, Images, Discover).
  • Narrow/gated: Course — 3+-course carousel only; Movie — mobile-only carousel.
  • Separate property: Dataset — feeds Dataset Search, not the normal SERP (Google clarified this Nov 2025).
  • Signal only: Article — better title/image/date display, no dedicated card, no AMP requirement.
  • No dedicated feature found: AudioObject (absent from Google’s current gallery — negative evidence limited to that gallery, not proof of zero use elsewhere) and Book (no mainstream card; narrow Book Actions feed program still live).
• June 2025 inflection: Course Info single-card announced for deprecation June 12, 2025 and removed (3+-course Course list carousel remains, a separate feature); Critic Review layer deprecated (Movie carousel survives, mobile-only, separately documented); Book Actions announced for deprecation June 2025 but its deprecation banner was removed November 5, 2025 — don’t say it’s gone.
• Article correction: no required properties, no dedicated visual card, and Article schema does not require AMP (stale/legacy advice).
• How to choose: pick the most accurate subtype first, then check consumer-specific eligibility. Recipe/Video → high-ROI, real results; Course/Movie/Dataset → narrow fit; Article/Audio/Book → entity understanding only, no verified AI-specific benefit, no special listing.
• Shared rules: required properties gate eligibility; eligibility ≠ display; mark up only visible/accurate content; JSON-LD recommended; not a ranking factor.

Official documentation

Primary-source documentation for the eight creative-works types and the rules they share. Note the pattern: Google publishes a per-type guide only where a rich result exists — AudioObject has none.

Google — the shared rules and the master list

• Structured data search gallery — the supported-types list; the source of truth for what actually produces a rich result. (Note which of these eight appear — and that AudioObject does not.)
• General structured data guidelines — the eligibility principle, content-quality and spam policies across all types.
• Simplifying our structured data support (June 2025) — the deprecation announcement (Course Info, Critic Review, Book Actions, and four others).

Google — Tier 1 (mainstream visual result)

• Recipe structured data — the mature recipe rich result; requires image and name.
• Video structured data — VideoObject across SERP, Video tab, Images, and Discover; requires name, thumbnailUrl, uploadDate.

Google — Tier 2/3 (narrow or separate-property)

• Course list structured data — the current 3+-course carousel (the single-course card was deprecated).
• Movie carousel structured data — the mobile-only movie carousel; requires image, name.
• Dataset structured data — feeds Dataset Search; requires description, name.

Google — Tier 4 (signal only)

• Article structured data — no required properties, no dedicated visual card, no AMP requirement.

Google — the one narrow place the base type appears

• Subscription and paywalled content structured data — lists CreativeWork (the base type) and several subtypes as supported page types for this bounded profile; this is the exception to “Google never uses CreativeWork.”

Schema.org (taxonomy)

• schema.org/CreativeWork — the formal umbrella type. Pure vocabulary; it never mentions Google, SEO, or rich results.

Bing / Microsoft

• Marking up your site with structured data — generic schema.org support; no per-type rich-result specs.

Quotes from the source

On-the-record statements from Google’s own documentation. Where a page exposes the text, the link is a deep link that jumps to the quoted passage.

Google docs — Tier 1: the two types with real visual results

• “Videos can appear in several different places on Google, including the main search results page, Video mode, Google Images, and Google Discover.” Jump to quote

Google docs — Tier 2: Course, the narrow feature

• “You must mark up at least three courses.” Jump to quote

Google docs — Tier 2: Movie is mobile-only

• “The movie carousel is only available on mobile devices.” Jump to quote

Google docs — Tier 3: Dataset lives in a separate tool

• “Datasets are easier to find in the Dataset Search tool when you provide supporting information such as their name, description, creator and distribution formats as structured data.” Jump to quote

Google docs — Tier 4: Article has no required properties (and no dedicated card)

• “There are no required properties; instead, add the properties that apply to your content.” Jump to quote

Patrick Stox — the thesis behind this whole hub

• “I’m a fan of schema markup as long as it gets you a search feature.” From my Enterprise SEO guide on Ahrefs

Creative Works schema at a glance

The eight types by rich-result tier

Minimum required properties (Tiers 1–3, where a result exists)

Fast facts

• “Creative Works schema” is a practitioner grouping for eight of the schema.org CreativeWork subtypes (not exhaustive) — schema.org groups them into one type; Google does not group them into one rich-result feature (though Google does use the base CreativeWork type narrowly, in its paywalled-content profile).
• Valid ≠ visible. Only Recipe and VideoObject earn a mainstream visual result. Prioritize those; the rest are narrow or understanding-only.
• June 2025: Course Info card deprecated; Critic Review deprecated; Book Actions announced but not removed.
• Article does not require AMP — that’s stale advice.
• Eligibility ≠ display, and schema is not a ranking factor — it earns a feature, not a rank boost.
• JSON-LD is Google’s recommended format for all eight.

Which creative-works type applies — and will it get me a visible result?

Answer for the content you’re marking up. This tree lands you on the right type and tells you honestly whether to expect a visible rich result or just vocabulary/entity understanding with no verified AI-specific benefit.

Anti-patterns — the mistakes this hub exists to correct

Assuming any of the eight always earns a visible rich result. This is the myth the whole hub is built to bust. “Add schema → get a fancy listing” is only true for Recipe and VideoObject. Adding Article, AudioObject, or Book schema and then waiting for a special SERP treatment that isn’t coming wastes the expectation, if not the effort — mark those up for vocabulary/entity understanding and general eligibility (not a verified AI-specific benefit), not for a visual result that doesn’t exist for that type.

Adding AMP because you think Article schema requires it. There’s no AMP requirement anywhere in Google’s current Article structured-data guide, and there are no required properties at all — just add the ones that apply to your content. This is stale, legacy advice, and I still see it repeated a lot. Don’t build AMP infrastructure to satisfy a rule Google doesn’t have.

Building a single-course rich-result card. The Course Info card (price, dates, instructor for one course) was deprecated on June 12, 2025. What remains is the Course list carousel, and it requires at least three courses marked up together. If you’re marking up one course expecting a dedicated card, you’re building for a feature that no longer exists.

Telling a client Book Actions was removed. It was in the June 2025 deprecation announcement batch, but it was the one type excluded from the actual September 2025 removal — Google later took the deprecation banner off the live Book doc. Book Actions is narrow and registration-gated, but it’s still live. Saying it’s gone is the opposite error from expecting a rich result that never existed.

Treating “Creative Works schema” as an official Google feature grouping. CreativeWork is a real schema.org umbrella type, and Google does directly recognize the base type in one narrow context (its paywalled-content structured-data profile) — but Google never groups these eight types into a “creative works” rich-result category in its own documentation; each type gets its own standalone guide, with zero cross-linking between them. Citing “Creative Works schema” as something Google rewards as one feature family, rather than my own practitioner grouping for related types, misstates the source.

Tools for validating and building creative-works schema

These two tools check different things, and passing either one is not the same as qualifying for a visible result — worth being precise about, since that conflation is exactly the mistake this hub is built to correct:

My Schema Markup Validator checks vocabulary structure — is this valid schema.org JSON-LD, correctly typed and nested. Paste the block (or a full page) and it runs severity-tiered checks against schema.org’s vocabulary and Google’s rich-result requirements — for example it catches a Recipe missing image or name, or a VideoObject missing thumbnailUrl or uploadDate, and hands back a corrected, copy-pasteable JSON-LD block. A clean pass here confirms the markup is well-formed; it does not confirm Google will display anything for it.

My Rich-Result Eligibility Checker checks Google-supported profiles — the question this whole hub is built around: not “is this valid JSON-LD” but “does this page actually qualify for a visible result.” Paste JSON-LD, an HTML page, or fetch a live URL, and it shows, per type, which required fields are present versus missing — useful for confirming, before you publish, whether you’re looking at a type with a real Google feature (Recipe, VideoObject) that’s actually eligible, or a type with no dedicated visual card (Article, AudioObject, Book) where no amount of extra markup will produce one.

If you’re starting from a blank page rather than fixing existing markup, my Schema Markup Generator has focused forms for building Recipe, VideoObject, Course, Movie, Dataset, and Article JSON-LD without hand-writing it, and it flags whether each property is Google-required, Google-recommended, or schema.org-only as you fill it in.

Once markup passes those checks, run the page through Google’s own Rich Results Test — it’s the tool Google actually uses to decide eligibility, so it’s the final word before you ship. But even a clean Rich Results Test pass confirms eligibility, not that Google will choose to display the enhancement — eligibility is never a display guarantee.

Test yourself: Creative Works Schema

Five quick questions on the eight creative-works types and — the point of this hub — which ones actually earn a visible result. Pick an answer for each, then check.

Resources worth your time

My schema markup writing

• Schema Markup — my broader take on the vocabulary, the formats, entity understanding, and the deprecation cycle. Creative works schema is a slice of this.
• Enterprise SEO — my pragmatic rule, which is exactly the thesis of this hub’s valid-vs-visible framing: “I’m a fan of schema markup as long as it gets you a search feature.”
• The Beginner’s Guide to Technical SEO — where I frame schema as code that helps engines understand content and powers the features that make a listing stand out.

From around the industry

The primary sources — trust these first:

• schema.org/CreativeWork — the formal umbrella type and its subtype hierarchy, straight from the vocabulary. Note it never mentions Google, SEO, or rich results — which is exactly why no single existing resource bridges “what’s valid” (schema.org) with “what’s rewarded” (Google).
• Structured data search gallery (Google Search Central) — the supported-types list; the honest source of truth for what actually earns a result.
• Simplifying our structured data support (Google Search Central) — the June 2025 deprecation announcement behind the Course Info, Critic Review, and Book Actions changes.

Cited honestly as examples of the conflation this hub exists to correct — not as endorsed sources:

• Yoast’s Article schema explainer is a clear example of the overstatement to watch for — it implies Article schema “can enable rich results like Top Stories, article carousels, and visual enhancements,” when Article has no dedicated visual card. I mention it only to illustrate the valid-vs-visible confusion, not to endorse it.
• In my reading, some Semrush and AIOSEO guidance blurs the distinction between adding Article or Book markup and qualifying for a Google rich result. I would treat those guides as implementation perspectives, not as the source of truth for current Google Search appearance eligibility.
• schemapros.com and schemaapp.com are among the very few that use a “Creative Work(s)” label at all — useful as precedent for the grouping, though schemapros.com references the deprecated Structured Data Testing Tool rather than the current Rich Results Test, an outdated-tool gap to avoid.

Creative Works Schema

"Creative Works schema" is a practitioner label for the schema.org CreativeWork subtypes used to mark up authored content — Article, AudioObject, Book, Course, Dataset, Movie, Recipe, and VideoObject.

Related: Article Schema, Recipe Schema, VideoObject Schema, Dataset Schema

Creative Works Schema

“Creative Works schema” groups the schema.org CreativeWork subtypes that describe authored content — Article, AudioObject, Book, Course, Dataset, Movie, Recipe, and VideoObject. It’s not an official Google or schema.org grouping; each subtype has its own eligibility rules, and only some (Recipe, VideoObject) reliably earn a visible rich result in Google Search.

Related: Article Schema, Recipe Schema, VideoObject Schema, Dataset Schema

Changelog