HowTo Schema

TL;DR — HowTo schemaHowTo schema (schema.org/HowTo) is structured data for step-by-step instructions. Google fully removed the HowTo rich result — on both desktop and mobile — in 2023, so it no longer produces any visible treatment in Search, though the schema.org vocabulary remains valid to use. is code you add to a page with step-by-step instructions — it labels “these are the steps,” “these are the tools you need,” “this is how long it takes.” It used to get you a special listing in Google with the steps shown right in the search results. That listing is gone — Google removed it completely in 2023, on both phones and desktop. The code is still valid to use and won’t hurt anything, but it no longer gets you anything extra in Google Search.

What HowTo schema is

When you write a “how to” article — how to tie a knot, how to reset a router, how to assemble a shelf — a person reading it can tell the steps apart, spot which tools they’ll need, and see roughly how long it’ll take. A search engine just sees text. HowTo schema spells all of that out in code, using the shared schema.org vocabulary, so the engine knows “this is step 1,” “this is a tool,” “this is the total time.”

It’s almost always written as JSON-LDJSON-LD (JavaScript Object Notation for Linked Data) is a script-based structured data format, typically paired with the schema.org vocabulary to describe page content for search engines and AI systems. Google recommends it over Microdata and RDFa because it's the easiest format to implement and maintain at scale — but all three work, and structured data isn't a ranking signal. — a small block of code that sits in the page without changing how the page looks.

The important part: the Google rich result is gone

For years, marking up a how-to page could earn you a rich result — an expandable set of step cards, sometimes with images, shown right in Google’s search results. That was the whole reason most people added it.

Google removed that rich result completely in 2023. It happened in two steps: first it disappeared from mobile (August 8, 2023), and then from desktop too (as of September 13, 2023). So today, no matter what device someone is on, adding HowTo schema gets you no special listing in Google Search.

A common mix-up worth clearing up right away: this was not a “desktop-only” change, and it isn’t still showing on some hidden surface. It’s a full removal, everywhere.

Do you need to remove old HowTo markup?

No. Google has said plainly that structured dataStructured data is a standardized way of labeling page content (using the schema.org vocabulary in JSON-LD, Microdata, or RDFa) so search engines can understand its meaning. It's not a direct ranking factor — its value is rich results and entity understanding. you’re no longer using doesn’t cause any problems — it just has no visible effect. So if you added HowTo schema years ago, you can leave it in place. There’s no penalty and no urgency to rip it out.

Don’t confuse it with Recipe schema

If you run a recipe or food site, the type you care about is Recipe schemaRecipe schema (schema.org/Recipe) marks up a cooking recipe's image, name, ratings, times, ingredients, and nutrition so the page can earn Google's recipe rich result — a visual card with star ratings and cook time, plus carousel and recipe-filter eligibility. — which is a different thing that still works and still earns rich resultsRich results (formerly 'rich snippets') are enhanced search listings — stars, images, prices, breadcrumbs, video thumbnails, and more — that Google and Bing build from structured data. They're a display feature, not a ranking factor, and eligibility never guarantees they'll show. (ratings, cook times, images). HowTo and Recipe often get lumped together, but only Recipe still pays off in Google’s results today.

Want the full version — the exact timeline, the schema.orgSchema markup is code that uses the schema.org vocabulary to label what your content means so search engines can understand it and show rich results. It's most often written in JSON-LD, and it's not a direct ranking factor. properties, what Google did to its own docs, and whether the markup has any leftover value for AI? Switch to the Advanced tab.

TL;DR — schema.orgSchema markup is code that uses the schema.org vocabulary to label what your content means so search engines can understand it and show rich results. It's most often written in JSON-LD, and it's not a direct ranking factor./HowTo marks up step-by-step instructions (step, HowToStep, HowToSection, tool, supply, totalTime, prepTime, performTime, estimatedCost, yield). Its Google rich result was removed in two stages — mobile on August 8, 2023, then desktop as of September 13, 2023 — a full removal Google’s own changelog describes as covering “both desktop and mobile devices.” No documented exception exists for Assistant or any region; treat “it still shows somewhere” as an unverified rumor. Google then removed the HowTo docs page itself (a stronger deprecation signal than FAQPage got). The schema.org vocabulary is unchanged and still valid, and Google says there’s no need to proactively remove old markup. Bing has no dedicated HowTo rich result either. Recipe schemaRecipe schema (schema.org/Recipe) marks up a cooking recipe's image, name, ratings, times, ingredients, and nutrition so the page can earn Google's recipe rich result — a visual card with star ratings and cook time, plus carousel and recipe-filter eligibility. is a separate type that still earns rich results — don’t conflate them. The only remaining upside — LLMA large language model (LLM) is a deep-learning model trained on massive text corpora to predict the next token and generate human-like text. LLMs use the transformer architecture and power AI search features like Google's AI Overviews (Gemini) and Bing Copilot (GPT-4)./AI-Overview parsing of structured step data — is speculative and unproven.

What HowTo schema is

HowTo is the schema.org type for instructional content: “Instructions that explain how to achieve a result by performing a sequence of steps.” You use it to describe the ordered steps of a task, plus the tools and supplies it needs and the time it takes. This nests under structured dataStructured data is a standardized way of labeling page content (using the schema.org vocabulary in JSON-LD, Microdata, or RDFa) so search engines can understand its meaning. It's not a direct ranking factor — its value is rich results and entity understanding. on this site because it’s the same schema.org/JSON-LDJSON-LD (JavaScript Object Notation for Linked Data) is a script-based structured data format, typically paired with the schema.org vocabulary to describe page content for search engines and AI systems. Google recommends it over Microdata and RDFa because it's the easiest format to implement and maintain at scale — but all three work, and structured data isn't a ranking signal. machinery as every other type — the difference is what Google does (or, here, no longer does) with it.

Is HowTo schema still worth using in 2026?

Short answer, up front, because it’s what you came for: as a Google rich-result play, no. The rich result doesn’t exist anymore. Adding HowTo markup earns you zero visible treatment in Google Search on any device. If your only goal was the step-card SERP featureSERP features are any element on a search results page beyond the classic ten blue links — featured snippets, People Also Ask, knowledge panels, sitelinks, image and video packs, AI Overviews, and structured-data-driven rich results. Google documents that losing rich-result eligibility doesn't affect ranking; their SEO relevance is CTR (they redistribute clicks). Some are unlocked by markup; most are purely algorithmic., there’s nothing left to chase.

The longer answer — whether it’s worth keeping for other reasons — is at the bottom of this section, and the honest version is “maybe, but nobody has proven it.”

A quick way to decide, depending on who you are:

• Already have HowTo markup on old pages? Leave it. Google says unused structured data causes no problems, so there’s no rush to strip it out.
• Building a new how-to page from scratch, chasing the rich result? Don’t bother adding HowTo just for that — the result Google used to show doesn’t exist anymore, on any device.
• Weighing whether to keep it for AI/LLM parsing? That’s the one open question below, and it’s a bet on an unproven upside, not a documented one. Adding valid markup is cheap if you’re already touching the template, but validating the markup doesn’t by itself create any of this value — it only confirms the vocabulary is well formed.
• Running a recipe or food site? Check Recipe schemaRecipe schema (schema.org/Recipe) marks up a cooking recipe's image, name, ratings, times, ingredients, and nutrition so the page can earn Google's recipe rich result — a visual card with star ratings and cook time, plus carousel and recipe-filter eligibility. instead — it’s the type that still earns a rich result.

TIP Verify the feature status before maintaining the markup

The checker can identify valid HowTo vocabulary and its current Google status as separate facts. A clean block does not restore a rich result Google removed.

Paste the deployed page or JSON-LD into my eligibility checker to confirm the detected type and the current feature record. Rich-Result Eligibility Checker Free

1. Confirm the visible instructions and detected HowTo markup still match.
2. Read the deprecated feature status separately from schema.org validity.
3. Keep or remove the markup based on maintenance value, not an expected Google rich result.

Tap to enlarge

The removal timeline: mobile first, then desktop — a full removal

This is the part I most want people to get right, because even this site’s own structured-data hub described it imprecisely, and a lot of stale guides get it backwards. Here’s exactly what happened, straight from Google’s announcement:

• August 8, 2023 — removed from mobile. In the original post, Google said it was “limiting How-To rich resultsRich results (formerly 'rich snippets') are enhanced search listings — stars, images, prices, breadcrumbs, video thumbnails, and more — that Google and Bing build from structured data. They're a display feature, not a ranking factor, and eligibility never guarantees they'll show. to desktop devices,” i.e. HowTo would “only be shown for desktop users, and not for users on mobile devices.” Mobile lost it first. (This is the opposite of the common myth that it was “removed from desktop but still shows on mobile.”)
• September 13, 2023 — removed from desktop too. A little over a month later, Google extended the change: “As of September 13, Google Search no longer shows How-to rich resultsRich results (formerly 'rich snippets') are enhanced search listings — stars, images, prices, breadcrumbs, video thumbnails, and more — that Google and Bing build from structured data. They're a display feature, not a ranking factor, and eligibility never guarantees they'll show. on desktop, which means this result type is now deprecated.” That’s the whole thing gone, on every device.

Google’s documentation changelog put the scope beyond argument: it describes removing the HowTo docs “as this rich result is no longer shown in search results, on both desktop and mobile devices.” “Both desktop and mobile devices” is the load-bearing phrase. There is no surface in Google Search where the HowTo rich result still renders.

The reporting side had an announced schedule too. Google said it would drop the How-to search appearance, rich result report, and Rich Results Test support within 30 days, and remove Search Console APIA set of REST APIs that let you programmatically read and manage Google Search Console data for properties you've verified — Performance data, URL index status, sitemaps, and properties — authorized with OAuth 2.0. support for How-to within 180 days. That’s Google’s stated plan (roughly October 2023 for the reports, roughly March 2024 for the API) — I haven’t independently re-verified that the reports and API support were actually pulled on those exact dates, so treat the dates as the announced schedule, not a confirmed completion log.

Google removed its own documentation page

Here’s how permanent this is. Google didn’t just stop renderingTurning HTML, CSS, and JavaScript into the final visual page and DOM. the result — it deleted the HowTo structured-data documentation page, replacing it with the changelog note above. The old /search/docs/appearance/structured-data/how-to URL now redirectsA redirect sends browsers and crawlers from a requested URL to a different one. An HTTP redirect specifically is a 3xx status code paired with a Location header; meta refresh and JavaScript redirects achieve a similar navigation without being a 3xx response themselves. Permanent redirects (301/308) are Google's signal the target should be canonical; temporary ones (302/303/307) aren't. to that deprecation entry. That’s a stronger deprecation signal than its twin got in the same announcement: FAQPage was reduced, not eliminated (its docs page still exists, just scoped down so FAQ rich results only show for “well-known, authoritative government and health websites”). HowTo got the harder cut.

Does it still show anywhere? (Assistant, other regions)

No, as far as any official documentation shows. The August post said the change “should finish rolling out globally,” and the September update extended removal to desktop with no geographic or platform carve-out. I’ve seen claims floating around that HowTo still surfaces in Google Assistant or in certain countries — I could not find anything official to support that, and Google’s global-rollout language plus the “both desktop and mobile” doc note point the other way. So I’ll say it the careful way: no publicly documented exception exists. If you hear “it still works in Assistant / some regions,” treat that as an unverified rumor, not fact.

HowTo schema properties and structure

The schema.org vocabulary itself was never touched — this was a Google Search product decision, not a schema.org spec change. The type and all its properties still exist and still validate. The direct properties of HowTo worth knowing:

• step — the ordered steps. Each is a HowToStep (“A step in the instructions for how to achieve a result”), or you can group steps into a HowToSection (“A sub-grouping of steps… e.g. steps for making a pie crust within a pie recipe”).
• tool — something used but not consumed while performing the steps (a screwdriver).
• supply — something consumed while performing the steps (glue, an egg).
• totalTime — total time to complete, in ISO 8601 duration format (PT1H30M).
• prepTime — time to prepare the supplies.
• performTime — time to actually do the steps, not counting prep.
• estimatedCost — the estimated cost of the supplies (a MonetaryAmount or text).
• yield — what you end up with (“a paper airplane, 10 personalized candles”). Evidence for this claim Schema.org HowTo supports steps, sections, supplies, tools, duration, and other instructional properties. Scope: Schema.org vocabulary; publishers should only mark up content actually represented on the page. Confidence: high · Verified: 2026-07-14 Schema.org: HowTo

HowToStep vs. HowToSection

HowTo.step is an array. In the simple case it’s a flat list of HowToStep items. When the steps naturally group — a “Prep” phase and an “Assembly” phase — you nest a HowToSection that itself holds an itemListElement array of HowToSteps. That grouping is the whole reason HowToSection exists. None of this earns you anything in Google Search anymore; it’s just how the vocabulary is structured if you’re using it for other reasons.

Should you remove HowTo markup from your site?

No — and this is official, not my opinion. On dropping this structured data, Google said: “While you can drop this structured data from your site, there’s no need to proactively remove it. Structured data that’s not being used does not cause problems for Search, but also has no visible effects in Google Search.” So leaving old HowTo blocks in place is harmless. Anyone telling you to urgently strip HowTo schemaHowTo schema (schema.org/HowTo) is structured data for step-by-step instructions. Google fully removed the HowTo rich result — on both desktop and mobile — in 2023, so it no longer produces any visible treatment in Search, though the schema.org vocabulary remains valid to use. “or it’ll hurt your SEO” is wrong.

If you’re doing a cleanup pass for maintainability reasons — fewer moving parts, less dead code to trip over — that’s a fine reason to remove it. Just don’t do it in a panic about penalties, because there aren’t any.

HowTo vs. Recipe vs. FAQPage

These get conflated constantly, so here’s the clean split:

• HowTo — generic step-by-step instructions. Rich result fully removed (2023, all devices). Vocabulary still valid.
• Recipe — a separate, still-supported type that still earns rich results (ratings, cook time, images) and is still promoted by Google. Recipe and HowTo can coexist on the same page (Recipe for the dish, a nested HowTo for a general technique), but only Recipe pays off in the SERP. Conflating the two overstates HowTo’s relevance.
• FAQPage — cut in the same 2023 announcement as HowTo, but only reduced: FAQ rich results now show only for well-known, authoritative government and health sites. Its docs page survives; HowTo’s didn’t.

Does HowTo schema still have value for AI Overviews / LLMs?

This is the one live argument for keeping the markup, and I want to be scrupulous about how strong it actually is: it’s a hypothesis, not a proven benefit.

The theory goes: structured, machine-readable step data might help LLMs and AI-Overview-style systems parse and possibly cite instructional content more reliably than they’d parse loose HTML. It’s plausible. It’s also unproven — I haven’t found a controlled study showing HowTo markup measurably improves AI citation or visibility, and I’m generally skeptical of “schema helps LLMs” claims without evidence to back them. So treat it as an emerging practice some practitioners bet on, not an established lever. Google does not require special Schema.org markup for its AI searchAI search uses large language models and retrieval-augmented generation (RAG) to synthesize an answer from multiple sources rather than returning a ranked list of links. Examples include Google AI Overviews, ChatGPT Search, and Perplexity. features. Evidence for this claim Google states that no special Schema.org structured data is required for its AI search features. Scope: Google Search AI-features guidance; HowTo markup is not a confirmed AI Overview or citation boost. Confidence: high · Verified: 2026-07-14 Google: AI features and your website If keeping the markup is cheap and already there, leaving it costs you nothing (per Google’s own guidance above). Just don’t build a business case on an unmeasured upside.

Bing and other search engines

Bing has no dedicated HowTo rich result and no HowTo-specific documentation. Its general structured-data guidance covers schema.org broadly and recommends JSON-LDJSON-LD (JavaScript Object Notation for Linked Data) is a script-based structured data format, typically paired with the schema.org vocabulary to describe page content for search engines and AI systems. Google recommends it over Microdata and RDFa because it's the easiest format to implement and maintain at scale — but all three work, and structured data isn't a ranking signal., but it doesn’t call out HowTo as a supported rich-result type. So there’s no Bing HowTo feature to gain by adding the markup — treat Bing as “general schema.org support, no HowTo rich result.”

For where HowTo sits in the wider picture, see the broader structured data and schema markupSchema markup is code that uses the schema.org vocabulary to label what your content means so search engines can understand it and show rich results. It's most often written in JSON-LD, and it's not a direct ranking factor. hubs this article nests under, and the sibling types in this cluster like FAQPage schemaFAQPage schema (schema.org/FAQPage) is structured data for a page's frequently-asked-questions content. It used to make pages eligible for an expandable FAQ rich result in Google Search, but Google fully removed that rich result on May 7, 2026 — the markup itself stays valid schema.org vocabulary, Google says it still reads it for understanding, and Bing has announced no equivalent removal. and Recipe schema.

AI summary

A condensed take on the Advanced version:

• What it is: schema.org/HowTo markup (usually JSON-LDJSON-LD (JavaScript Object Notation for Linked Data) is a script-based structured data format, typically paired with the schema.org vocabulary to describe page content for search engines and AI systems. Google recommends it over Microdata and RDFa because it's the easiest format to implement and maintain at scale — but all three work, and structured data isn't a ranking signal.) that labels step-by-step instructions — step/HowToStep/HowToSection, tool, supply, totalTime, prepTime, performTime, estimatedCost, yield.
• The rich result is gone — in two stages: removed from mobile first (August 8, 2023), then from desktop too (as of September 13, 2023). Google’s documentation changelog says the removal covers “both desktop and mobile devices.” It is not a desktop-only change — that’s a common myth, and it’s backwards (mobile went first).
• No documented exception for Google Assistant or any region. “It still shows somewhere” is an unverified rumor, not fact.
• Google removed its own HowTo docs page (redirectsA redirect sends browsers and crawlers from a requested URL to a different one. An HTTP redirect specifically is a 3xx status code paired with a Location header; meta refresh and JavaScript redirects achieve a similar navigation without being a 3xx response themselves. Permanent redirects (301/308) are Google's signal the target should be canonical; temporary ones (302/303/307) aren't. to the changelog) — a harder cut than FAQPage, which was only reduced (gov/health sites only) and kept its docs.
• Reporting wind-down was an announced schedule: Google said How-to reports/Rich ResultsRich results (formerly 'rich snippets') are enhanced search listings — stars, images, prices, breadcrumbs, video thumbnails, and more — that Google and Bing build from structured data. They're a display feature, not a ranking factor, and eligibility never guarantees they'll show. Test support would drop ~30 days after, and Search Console API support ~180 days after — stated timelines, not independently confirmed completion dates.
• Keep or remove? Google says there’s no need to proactively remove unused structured dataStructured data is a standardized way of labeling page content (using the schema.org vocabulary in JSON-LD, Microdata, or RDFa) so search engines can understand its meaning. It's not a direct ranking factor — its value is rich results and entity understanding. — it causes no problems and has no visible effect. No penalty for leaving it.
• Recipe schemaRecipe schema (schema.org/Recipe) marks up a cooking recipe's image, name, ratings, times, ingredients, and nutrition so the page can earn Google's recipe rich result — a visual card with star ratings and cook time, plus carousel and recipe-filter eligibility. is separate and still works (real rich resultsRich results (formerly 'rich snippets') are enhanced search listings — stars, images, prices, breadcrumbs, video thumbnails, and more — that Google and Bing build from structured data. They're a display feature, not a ranking factor, and eligibility never guarantees they'll show.). Don’t conflate HowTo with Recipe.
• Bing: no dedicated HowTo rich result; general schema.orgSchema markup is code that uses the schema.org vocabulary to label what your content means so search engines can understand it and show rich results. It's most often written in JSON-LD, and it's not a direct ranking factor. support only.
• AI/LLMA large language model (LLM) is a deep-learning model trained on massive text corpora to predict the next token and generate human-like text. LLMs use the transformer architecture and power AI search features like Google's AI Overviews (Gemini) and Bing Copilot (GPT-4). value: the idea that structured step data helps AI OverviewsAI Overviews are the AI-generated summary box Google shows above or within its regular search results, written by Gemini models from pages retrieved out of Google's normal Search index. It's a Search feature, not a separate platform or index./LLMs parse instructions is speculative and unproven — an emerging bet, not a confirmed benefit.

Official documentation

Primary-source documentation and the schema.orgSchema markup is code that uses the schema.org vocabulary to label what your content means so search engines can understand it and show rich results. It's most often written in JSON-LD, and it's not a direct ranking factor. spec.

Google

• Changes to HowTo and FAQ rich results (Search Central Blog, Aug 8, 2023, updated Sept 14, 2023) — the announcement and the September update banner extending removal to desktop.
• Search Central documentation updates — HowTo deprecation — the changelog entry stating the removal covers “both desktop and mobile devices” (this is what the old HowTo docs URL now redirectsA redirect sends browsers and crawlers from a requested URL to a different one. An HTTP redirect specifically is a 3xx status code paired with a Location header; meta refresh and JavaScript redirects achieve a similar navigation without being a 3xx response themselves. Permanent redirects (301/308) are Google's signal the target should be canonical; temporary ones (302/303/307) aren't. to).

schema.org (the vocabulary — still valid)

• schema.org/HowTo — the type and its full property table.
• schema.org/HowToStep — a single step.
• schema.org/HowToSection — grouping steps into sections.

Bing / Microsoft

• Marking up your site with structured data (Bing Webmaster ToolsMicrosoft's free portal for monitoring and improving how a site appears in Bing search — the peer to Google Search Console, plus IndexNow instant indexing, richer backlink data, and keyword volumes. Because Bing's index also feeds Microsoft Copilot, it doubles as a window into AI-search visibility. help) — Bing’s general schema.org support; no dedicated HowTo rich result is documented.

Quotes from the source

On-the-record statements from Google’s official blog and changelog, plus the schema.orgSchema markup is code that uses the schema.org vocabulary to label what your content means so search engines can understand it and show rich results. It's most often written in JSON-LD, and it's not a direct ranking factor. definitions. Each link jumps to the quoted passage on the source page.

Google — the original August 8, 2023 change (mobile removal)

• “How-To (from HowTo structured dataStructured data is a standardized way of labeling page content (using the schema.org vocabulary in JSON-LD, Microdata, or RDFa) so search engines can understand its meaning. It's not a direct ranking factor — its value is rich results and entity understanding.) rich resultsRich results (formerly 'rich snippets') are enhanced search listings — stars, images, prices, breadcrumbs, video thumbnails, and more — that Google and Bing build from structured data. They're a display feature, not a ranking factor, and eligibility never guarantees they'll show. will only be shown for desktop users, and not for users on mobile devices.” — Google Search Central Blog, “Changes to HowTo and FAQ rich resultsRich results (formerly 'rich snippets') are enhanced search listings — stars, images, prices, breadcrumbs, video thumbnails, and more — that Google and Bing build from structured data. They're a display feature, not a ranking factor, and eligibility never guarantees they'll show..” Jump to quote

Google — the September 14, 2023 update (desktop removal = full removal)

• “As of September 13, Google Search no longer shows How-to rich results on desktop, which means this result type is now deprecated.” — Google Search Central Blog, update banner. Jump to quote
• “Since How-to results no longer appear in Google Search, we will be dropping the How-to search appearance, rich result report, and support in the Rich results test in 30 days. To allow time for adjusting your API calls, support for How-to in the Search Console APIA set of REST APIs that let you programmatically read and manage Google Search Console data for properties you've verified — Performance data, URL index status, sitemaps, and properties — authorized with OAuth 2.0. will be removed in 180 days.” — Google Search Central Blog, update banner. Jump to quote

Google — the documentation changelog (full scope, both devices)

• “Removed the How-to structured data documentation, as this rich result is no longer shown in search results, on both desktop and mobile devices.” — Google Search Central, documentation updates changelog. Jump to quote

Google — on removing old markup

• “While you can drop this structured data from your site, there’s no need to proactively remove it. Structured data that’s not being used does not cause problems for Search, but also has no visible effects in Google Search.” — Google Search Central Blog, “Changes to HowTo and FAQ rich results.” Jump to quote

schema.org — the vocabulary (unchanged)

• HowTo: “Instructions that explain how to achieve a result by performing a sequence of steps.” Jump to definition
• HowToSection: “A sub-grouping of steps in the instructions for how to achieve a result (e.g. steps for making a pie crust within a pie recipe).” Jump to definition

HowTo schema cheat sheet

Status in Google Search (2026)

The removal timeline

Core schema.org/HowTo properties (still valid)

step (holds HowToStep, or grouped in HowToSection) · tool (used, not consumed) · supply (consumed) · totalTime · prepTime · performTime · estimatedCost · yield

Don’t confuse these

HowTo schema myths and mistakes to avoid

Myth: “HowTo was only removed from desktop; it still shows on mobile.” It’s the reverse, and then some. Mobile lost it first (August 8, 2023), then desktop lost it too (September 13, 2023). Nothing shows anywhere in Google Search now. Google’s changelog explicitly says “both desktop and mobile devices.”

Myth: “HowTo still works in Google Assistant / in certain countries.” No official documentation supports any surviving surface. Google’s language was a global rollout, and the doc-removal note says both devices. If you repeat this, frame it as an unverified rumor — not a fact.

Myth: “schema.orgSchema markup is code that uses the schema.org vocabulary to label what your content means so search engines can understand it and show rich results. It's most often written in JSON-LD, and it's not a direct ranking factor. deprecated HowTo, so you shouldn’t use the type.” schema.org did not deprecate it. Google stopped renderingTurning HTML, CSS, and JavaScript into the final visual page and DOM. a rich result for it — that’s a Google Search product decision. The vocabulary is still valid and still validates.

Myth: “You need to strip HowTo schemaHowTo schema (schema.org/HowTo) is structured data for step-by-step instructions. Google fully removed the HowTo rich result — on both desktop and mobile — in 2023, so it no longer produces any visible treatment in Search, though the schema.org vocabulary remains valid to use. off your pages now or it’ll hurt your SEO.” False. Google says unused structured dataStructured data is a standardized way of labeling page content (using the schema.org vocabulary in JSON-LD, Microdata, or RDFa) so search engines can understand its meaning. It's not a direct ranking factor — its value is rich results and entity understanding. “does not cause problems for Search” — there’s no penalty and no urgency. Remove it for cleanliness if you like, not out of fear.

Myth: “HowTo and Recipe schemaRecipe schema (schema.org/Recipe) marks up a cooking recipe's image, name, ratings, times, ingredients, and nutrition so the page can earn Google's recipe rich result — a visual card with star ratings and cook time, plus carousel and recipe-filter eligibility. are the same thing.” They’re separate types. Recipe schemaRecipe schema (schema.org/Recipe) marks up a cooking recipe's image, name, ratings, times, ingredients, and nutrition so the page can earn Google's recipe rich result — a visual card with star ratings and cook time, plus carousel and recipe-filter eligibility. still earns rich resultsRich results (formerly 'rich snippets') are enhanced search listings — stars, images, prices, breadcrumbs, video thumbnails, and more — that Google and Bing build from structured data. They're a display feature, not a ranking factor, and eligibility never guarantees they'll show. and is still promoted by Google; HowTo does not. Conflating them overstates HowTo’s relevance — if you run a recipe site, Recipe is the type that still matters.

Myth: “Adding HowTo markup helps AI OverviewsAI Overviews are the AI-generated summary box Google shows above or within its regular search results, written by Gemini models from pages retrieved out of Google's normal Search index. It's a Search feature, not a separate platform or index. cite my how-to content.” Maybe — but it’s unproven. There’s no controlled study showing HowTo markup improves AI citationAn AI citation is the visible source link an AI answer engine shows next to its generated text — the clickable reference that credits the web page it used. A citation's presence is a separate thing from whether the cited page actually supports the statement, and from being retrieved (read behind the scenes) or merely mentioned (named without a link); citation is driven more by brand mentions and being retrievable than by traditional ranking.. Treat it as a hypothesis, not a confirmed benefit, and don’t sell a client on an upside nobody has measured.

Mistake: following a stale implementation guide that leads with “get a rich snippet!” A lot of “HowTo schema” content online still hasn’t been updated since 2022 and implies the rich result is a live incentive. It isn’t. Check the date on any guide before you follow it.

What HowTo markup looks like (and what it earns you)

A valid HowTo block

This is a simplified but structurally correct HowTo, with tools, supplies, a total time, and ordered steps:

{
  "@context": "https://schema.org/",
  "@type": "HowTo",
  "name": "How to tie a bowline knot",
  "totalTime": "PT2M",
  "estimatedCost": {
    "@type": "MonetaryAmount",
    "currency": "USD",
    "value": "0"
  },
  "tool": [
    { "@type": "HowToTool", "name": "A length of rope" }
  ],
  "supply": [],
  "step": [
    {
      "@type": "HowToStep",
      "name": "Make a loop",
      "text": "Form a small loop near the end of the rope, leaving a long tail."
    },
    {
      "@type": "HowToStep",
      "name": "Pass the tail",
      "text": "Bring the tail up through the loop, around the standing line, and back down through the loop."
    },
    {
      "@type": "HowToStep",
      "name": "Tighten",
      "text": "Pull the standing line to snug the knot."
    }
  ]
}

This validates fine against schema.orgSchema markup is code that uses the schema.org vocabulary to label what your content means so search engines can understand it and show rich results. It's most often written in JSON-LD, and it's not a direct ranking factor.. In Google Search, it earns you nothing visible — the rich result it used to trigger no longer exists on any device.

Grouping steps with HowToSection

When steps naturally split into phases, nest them in a HowToSection:

{
  "@context": "https://schema.org/",
  "@type": "HowTo",
  "name": "How to build a birdhouse",
  "step": [
    {
      "@type": "HowToSection",
      "name": "Prep",
      "itemListElement": [
        { "@type": "HowToStep", "text": "Cut the panels to size." },
        { "@type": "HowToStep", "text": "Sand every edge." }
      ]
    },
    {
      "@type": "HowToSection",
      "name": "Assembly",
      "itemListElement": [
        { "@type": "HowToStep", "text": "Glue and clamp the sides." },
        { "@type": "HowToStep", "text": "Attach the roof." }
      ]
    }
  ]
}

The HowToSection grouping is the vocabulary working as designed — it’s just no longer surfaced by Google.

The visible content matters more than the markup

Since the markup itself earns nothing in Google Search, the value has to come from the actual page content being genuinely useful — to readers and to anything (a search engine, an LLMA large language model (LLM) is a deep-learning model trained on massive text corpora to predict the next token and generate human-like text. LLMs use the transformer architecture and power AI search features like Google's AI Overviews (Gemini) and Bing Copilot (GPT-4).) trying to parse it. A visible how-to page worth marking up, with or without the schema, generally covers:

• Ordered steps written as steps, not buried in a paragraph.
• Prerequisites — what the reader needs before starting (tools, supplies, skill level, time).
• Failure states — what it looks like when a step goes wrong, and how to recover.
• The outcome — what “done” looks like, so the reader can confirm success.

That’s a content-quality checklist, not a markup requirement. Marking up a thin or vague page doesn’t fix the page; the JSON-LDJSON-LD (JavaScript Object Notation for Linked Data) is a script-based structured data format, typically paired with the schema.org vocabulary to describe page content for search engines and AI systems. Google recommends it over Microdata and RDFa because it's the easiest format to implement and maintain at scale — but all three work, and structured data isn't a ranking signal. just describes whatever is already there.

HowTo vs. Recipe on the same page

A food page might carry both: Recipe for the dish (which still earns a rich result) and a nested or separate HowTo for a general technique (which does not). Keep them distinct — don’t assume the HowTo half is doing anything in the SERP just because the Recipe half is.

Tools for HowTo markup

• Schema Markup Validator checks whether HowTo, HowToStep, supplies, tools, and durations follow schema.orgSchema markup is code that uses the schema.org vocabulary to label what your content means so search engines can understand it and show rich results. It's most often written in JSON-LD, and it's not a direct ranking factor. vocabulary.
• Structured Data Generator provides a starting graph that must be edited to match the visible instructions.
• A crawlerA crawler — also called a spider or bot — is an automated program that fetches web pages, extracts their links, and queues new URLs to visit. Search engines use crawlers to discover and download content for their index. with JSON-LDJSON-LD (JavaScript Object Notation for Linked Data) is a script-based structured data format, typically paired with the schema.org vocabulary to describe page content for search engines and AI systems. Google recommends it over Microdata and RDFa because it's the easiest format to implement and maintain at scale — but all three work, and structured data isn't a ranking signal. extraction finds old HowTo markup, template gaps, and mismatches at scale.
• Google’s current rich-result tooling should not be interpreted as promising a HowTo search treatment: Google removed that rich result in 2023.

Validation can prove that the vocabulary is well formed and matches the page. It cannot establish a current Google rich-result benefit that no longer exists.

Test yourself: HowTo Schema

Five quick questions on schema.orgSchema markup is code that uses the schema.org vocabulary to label what your content means so search engines can understand it and show rich results. It's most often written in JSON-LD, and it's not a direct ranking factor./HowTo and the 2023 removal. Pick an answer for each, then check.

Resources worth your time

My writing on this topic

I haven’t published a standalone HowTo schemaHowTo schema (schema.org/HowTo) is structured data for step-by-step instructions. Google fully removed the HowTo rich result — on both desktop and mobile — in 2023, so it no longer produces any visible treatment in Search, though the schema.org vocabulary remains valid to use. article of my own — so rather than point you at something that doesn’t exist, I’ll send you to the primary sources and to the related structured-data work on this site. For the broader vocabulary and where HowTo fits, see the Structured DataStructured data is a standardized way of labeling page content (using the schema.org vocabulary in JSON-LD, Microdata, or RDFa) so search engines can understand its meaning. It's not a direct ranking factor — its value is rich results and entity understanding. and Schema MarkupSchema markup is code that uses the schema.org vocabulary to label what your content means so search engines can understand it and show rich results. It's most often written in JSON-LD, and it's not a direct ranking factor. hubs; for the AI angle (and my general skepticism about “schema helps LLMsA large language model (LLM) is a deep-learning model trained on massive text corpora to predict the next token and generate human-like text. LLMs use the transformer architecture and power AI search features like Google's AI Overviews (Gemini) and Bing Copilot (GPT-4).” claims without evidence), Schema Markup for AISchema markup (structured data) is machine-readable code — usually JSON-LD — that labels what your content means using the schema.org vocabulary. For AI search it's infrastructure for entity disambiguation, not a direct citation lever: controlled studies found no meaningful uplift in AI citations from adding it..

From around the industry

• Changes to HowTo and FAQ rich results (Google Search Central Blog) — the authoritative announcement, including the September desktop-removal update.
• Search Central documentation updates (Google) — the changelog note confirming removal on “both desktop and mobile devices.”
• schema.org/HowTo — the still-valid vocabulary and full property list.
• Google Search stops showing How-to rich results on desktop (Search Engine Land, Danny Goodwin) — coverage of the September 13 desktop cutoff and the reporting timeline.
• Google Completely Removes How-To Rich Results (Search Engine Journal, Roger Montti) — frames it as complete removal from both mobile and desktop, globally.
• Google Drops How-To Rich Results For Desktop Search (Search Engine Roundtable) — the practical site-owner framing (expect less traffic / lower CTR for pages that relied on the feature).

HowTo Schema

HowTo schema (schema.org/HowTo) is structured data for step-by-step instructions. Google fully removed the HowTo rich result — on both desktop and mobile — in 2023, so it no longer produces any visible treatment in Search, though the schema.org vocabulary remains valid to use.

Related: Structured Data, Schema Markup, Rich Results

HowTo Schema

HowTo schema is the schema.orgSchema markup is code that uses the schema.org vocabulary to label what your content means so search engines can understand it and show rich results. It's most often written in JSON-LD, and it's not a direct ranking factor./HowTo structured-data type used to mark up step-by-step instructional content — the ordered steps, plus the tools, supplies, and time required to complete a task. Its core properties include step (holding HowToStep items, optionally grouped in a HowToSection), tool, supply, totalTime, prepTime, performTime, estimatedCost, and yield.

It once powered a Google rich result — expandable step cards, sometimes with images, shown directly in the search results. That rich result is gone. Google removed it from mobile first (August 8, 2023) and then from desktop too (as of September 13, 2023), a full removal across both desktop and mobile devices. Google later pulled the HowTo documentation page itself, replacing it with a changelog note.

The markup is still valid schema.org vocabulary and Google says there’s no need to proactively remove it — it just produces no visible rich-result treatment in Google Search anymore. This is different from Recipe schemaRecipe schema (schema.org/Recipe) marks up a cooking recipe's image, name, ratings, times, ingredients, and nutrition so the page can earn Google's recipe rich result — a visual card with star ratings and cook time, plus carousel and recipe-filter eligibility., a separate type that still earns real rich resultsRich results (formerly 'rich snippets') are enhanced search listings — stars, images, prices, breadcrumbs, video thumbnails, and more — that Google and Bing build from structured data. They're a display feature, not a ranking factor, and eligibility never guarantees they'll show.. Any remaining value of HowTo markup (for example, helping LLMsA large language model (LLM) is a deep-learning model trained on massive text corpora to predict the next token and generate human-like text. LLMs use the transformer architecture and power AI search features like Google's AI Overviews (Gemini) and Bing Copilot (GPT-4). or AI systems parse instructional content) is speculative and unproven.

Related: Structured Data, Schema Markup, Rich Results

Changelog