NotionCue
AI Visibility Platform
All systems live
Sign in →
AEO Guidellms.txt GeneratorRobots.txtBLUF TemplatesBlogChangelogAbout
← Blog
TechnicalJun 29, 2026·10 min read

HowTo Schema for AEO: Why Step-by-Step Content Gets Cited Differently

Triple-schema stacking, FAQPage plus Article plus HowTo on one page, produces 1.8x more AI citations than Article schema alone. HowTo schema makes each step independently extractable. Here is the complete implementation with copy-paste JSON-LD.

SS
Sudhir Singh
Senior SEO & AEO Specialist · NotionCue
📋

Pages with three or more schema types are 13% more likely to earn AI citations than pages with no schema at all, per AirOps 2026 research. The highest-leverage combination is FAQPage plus Article plus HowTo, what practitioners call triple-schema stacking. This combination produces 1.8 times more citations than Article schema alone.

HowTo schema is the piece most teams skip. FAQPage and Article are well-documented. HowTo is treated as niche, useful only for recipe sites or assembly instructions. That instinct is wrong. Any page explaining a process, "how to set up prompt tracking," "how to audit AI crawler access," "how to fix dateModified schema", is procedural content. HowTo schema makes each step in that process independently extractable as a citable unit.

Without HowTo schema, an AI engine retrieving a process guide has to infer the step structure from your prose. It might get it right. It often does not. With HowTo schema, each step is machine-labelled, sequenced, and extractable without inference. The AI does not have to guess where Step 1 ends and Step 2 begins.

What Is HowTo Schema and How Does It Work?

HowTo schema is a schema.org type that structures procedural content as an ordered sequence of steps. Each step has a name, a description, and optionally an image. The structured step sequence is what AI engines use when generating step-by-step answer blocks, each HowToStep is a separately citable unit.

Google's HowTo rich results were deprecated for most query types in January 2026. This causes confusion. The deprecation removed HowTo from standard SERP rich results. It did not remove HowTo from AI retrieval. Google explicitly confirmed that structured data continues to support AI Overviews and AI Mode even after rich result deprecation. The same is true of FAQPage, deprecated from rich results in 2023 but still central to AI citation selection.

HowTo schema and FAQPage schema serve different extraction needs. FAQPage provides question-answer pairs, best for informational queries where the user wants a direct factual answer. HowTo provides ordered steps, best for procedural queries where the user wants to do something. Most process guides need both. A page explaining how to implement AEO schema should answer the "what is this" question (FAQPage) and explain the implementation steps (HowTo) in the same piece.

What JSON-LD Does a HowTo Page Need?

The minimum viable HowTo JSON-LD includes a name, a description, and at least two HowToStep objects. Each step needs a name (a short action title) and text (the step explanation in 40 to 60 words). Images per step are optional but improve extraction reliability in voice and visual AI surfaces.

{
  "@context": "https://schema.org",
  "@type": "HowTo",
  "name": "How to Implement FAQPage Schema for AEO",
  "description": "A step-by-step process for adding FAQPage JSON-LD to any page to improve AI citation rates across ChatGPT, Perplexity, and Google AI Overviews.",
  "totalTime": "PT30M",
  "estimatedCost": {
    "@type": "MonetaryAmount",
    "currency": "USD",
    "value": "0"
  },
  "step": [
    {
      "@type": "HowToStep",
      "position": "1",
      "name": "Identify pages with question-and-answer content",
      "text": "Review your top ten highest-traffic pages and list those that include at least three question-and-answer sections. These are your FAQPage schema candidates. Pages without any Q&A content need structural edits before schema can be added."
    },
    {
      "@type": "HowToStep",
      "position": "2",
      "name": "Write your FAQ questions in natural spoken language",
      "text": "Phrase each question exactly as a buyer would ask an AI engine. 'What is the best way to track AI citations?' earns citations. 'AEO tracking methodology overview' does not. Aim for five to eight questions per page, each with a 40 to 60 word direct answer."
    },
    {
      "@type": "HowToStep",
      "position": "3",
      "name": "Build the JSON-LD block",
      "text": "Create a script block with type application/ld+json. Add the FAQPage context, type, and mainEntity array containing your Question and Answer objects. Paste it into the head element of your page, or use your CMS schema plugin to insert it."
    },
    {
      "@type": "HowToStep",
      "position": "4",
      "name": "Validate and submit for re-crawl",
      "text": "Run the page through Google's Rich Results Test at search.google.com/test/rich-results. Fix any errors flagged. Submit the updated URL via Google Search Console URL Inspection tool and request indexing. Perplexity typically reflects schema changes within days."
    }
  ]
}

The totalTime field uses ISO 8601 duration format. PT30M means 30 minutes. This field is optional for AEO purposes but useful for voice assistants that read step-by-step content aloud and need to communicate estimated effort.

How Do You Stack HowTo With FAQPage and Article Schema?

Triple-schema stacking uses the JSON-LD @graph array to link multiple schema types on a single page as related entities. This is cleaner than three separate script blocks and allows entity relationships, the HowTo linking to the Article, the Article linking to the Person author, to be expressed explicitly.

{
  "@context": "https://schema.org",
  "@graph": [
    {
      "@type": "Article",
      "@id": "https://notioncue.com/blog/howto-schema-aeo/#article",
      "headline": "HowTo Schema for AEO: Step-by-Step Citation Guide",
      "datePublished": "2026-06-29",
      "dateModified": "2026-06-29",
      "author": {
        "@type": "Person",
        "@id": "https://notioncue.com/about/#person",
        "name": "Sudhir Singh"
      },
      "publisher": {
        "@type": "Organization",
        "@id": "https://notioncue.com/#organization"
      }
    },
    {
      "@type": "HowTo",
      "@id": "https://notioncue.com/blog/howto-schema-aeo/#howto",
      "name": "How to Implement HowTo Schema for AEO",
      "isPartOf": {
        "@id": "https://notioncue.com/blog/howto-schema-aeo/#article"
      },
      "step": [
        {
          "@type": "HowToStep",
          "position": "1",
          "name": "Identify procedural content pages",
          "text": "List every page that explains a process or workflow. Implementation guides, setup tutorials, and troubleshooting posts all qualify. AEO content gap analysis, schema setup, and prompt tracking configuration are common AEO-specific candidates."
        }
      ]
    },
    {
      "@type": "FAQPage",
      "@id": "https://notioncue.com/blog/howto-schema-aeo/#faq",
      "mainEntity": [
        {
          "@type": "Question",
          "name": "Does HowTo schema still work after Google deprecated rich results?",
          "acceptedAnswer": {
            "@type": "Answer",
            "text": "Yes. Google deprecated HowTo rich results in search listings in January 2026 but confirmed that structured data continues to support AI Overviews and AI Mode. HowTo schema still helps AI engines extract and cite step-by-step content. Do not remove it."
          }
        }
      ]
    }
  ]
}

The @id links use fragment identifiers to create internal relationships between the schema types. The isPartOf relationship on the HowTo connects it to the Article, telling AI engines that the step content is part of the broader editorial piece rather than a standalone procedure.

What Step Content Earns the Highest Citation Rates?

Not all HowToStep content is equally extractable. Three patterns produce the highest citation rates based on observed AI citation behaviour across procedural content.

Action-first step names. The name field in each HowToStep is the step headline an AI engine uses when generating a list answer. "Step 1: Identify pages" is weak. "Identify your top ten highest-traffic pages with question-and-answer content" is an action statement that carries meaning without the numbered label. Write step names as you would write a task on a to-do list: verb first, specific object second.

Self-contained step descriptions. Each step description should make sense without reading the other steps. An AI engine extracting only Step 3 needs to understand what Step 3 does. "Add the JSON-LD block to your page head element using the FAQPage schema format" is self-contained. "Now add it to your page" is not, "it" requires knowing what Step 2 said.

Specific not generic instructions. "Validate your schema" is a generic instruction. "Run your page URL through search.google.com/test/rich-results and fix any red errors before submitting for re-crawl" is specific. The specificity is what makes a step citable, generic instructions exist on thousands of pages; specific instructions with named tools and specific actions are more unique and therefore more citable.

What Are the Most Common HowTo Schema Errors?

Schema that fails silently is worse than no schema. A broken schema block is typically ignored by AI crawlers rather than partially processed, so the page receives zero schema benefit. Three errors cause most HowTo failures.

JavaScript-rendered schema. If your CMS outputs HowTo schema as JavaScript that renders after page load, AI crawlers may not see it. Run curl -A "Googlebot" https://yourpage.com and confirm the JSON-LD appears in the raw HTML response. If it does not, your schema plugin or theme is rendering it client-side. Fix this by ensuring schema is in server-rendered HTML.

Step text longer than 200 words. Long step descriptions reduce extraction reliability. An AI engine retrieving a step wants a concise action instruction, not a paragraph. Keep each step description between 40 and 150 words. If a step needs more explanation, split it into two steps.

HowTo name that does not match the page H1. The HowTo name field should match or closely mirror your page's H1 heading. A mismatch between schema name and visible page title creates a content-schema inconsistency that reduces citation confidence. AI engines cross-check schema claims against visible content. When they do not match, the schema loses weight.

The NotionCue AI Crawler Audit checks whether your schema is present in the initial HTML response or only in JavaScript-rendered DOM. Schema that lives in a client-side script is invisible to PerplexityBot and OAI-SearchBot. The audit surfaces this specifically, pages where schema is declared but not crawler-accessible are flagged separately from pages with missing schema entirely, so you know whether to fix the content or the rendering pipeline.

Frequently Asked Questions

Should every blog post have HowTo schema?
No. HowTo schema fits procedural content, pages that explain how to do something with sequential steps. An opinion piece, a case study, or a data analysis post does not have natural HowTo structure and should not have HowTo schema forced onto it. The right schema is the one that describes what the page actually contains. Applying HowTo schema to non-procedural content creates a schema-content mismatch that AI engines notice and penalise with reduced citation confidence.

How many steps should a HowTo schema have?
Three to eight steps for most process guides. Below three, the HowTo structure is barely different from a numbered list. Above eight, you are probably describing a complex process that benefits from being split into multiple pages within a content cluster, with each page targeting a more specific sub-process.

Can HowTo schema be added to a page that already has FAQPage schema?
Yes, and this combination is recommended for any guide that includes both a process and a Q&A section. Use the @graph stacking method shown above to link them as related entities on the same page. The 1.8x citation lift from triple-schema stacking comes specifically from this combination on a single well-structured page.

Does HowTo schema affect voice search as well as text AI citations?
Yes. Voice assistants, especially Google Assistant, retrieve HowTo schema to answer "how do I..." queries spoken aloud. Each step is read sequentially. This is why step names need to be action-first and self-contained: a voice listener cannot see the step number or re-read the previous step for context. For voice-specific optimisation on the same page, add Speakable schema alongside HowTo as covered in the Speakable schema guide.

Share this post
Check your AEO score
Scan your domain free — get your AI visibility score across 5 LLMs in 30 seconds.
Scan my site →
SS
Sudhir Singh
Senior SEO & AEO Specialist · NotionCue

Senior SEO and AEO specialist with 12+ years across e-commerce, global education, and healthcare. Building Notion Cue to track brand citations across ChatGPT, Perplexity, Gemini, and AI Overviews.

View all →
📖
Technical11 min read

The AEO Glossary: 25 Terms Every AI Search Practitioner Should Know

This is the first entry in a new reference series. Rather than another explainer, this is a working glossary — short, direct definitions for the terms that come up constantly across AI search, AEO, and GEO discussions, each linked to a deeper guide where one exists on this blog.

SS
Jul 27, 2026
🗺️
Technical9 min read

Sitemap Strategy for AI Crawlers: The Technical Layer Most AEO Audits Skip

Server log analysis across a meaningful sample of sites found GPTBot referencing sitemaps on roughly three out of four initial domain visits, compared to just over half for Googlebot. PerplexityBot showed an even stronger dependency, crawling sitemap-listed URLs several times more often than equivalent pages it found only through links. Most sitemaps are still built entirely around Google.

SS
Jul 22, 2026
📅
Technical8 min read

Event Schema for AEO: Why Your Conference or Webinar Needs Structured Data Now

Google's own list of schema types still worth investing in after its 2026 rich-result cleanup names Event alongside JobPosting as one of the categories that maps to genuine, ongoing business use rather than SERP-gaming. Most companies running webinars and conferences have never implemented it.

SS
Jul 20, 2026
🕷️
Technical7 min read

Google-CloudVertexBot and the New AI Crawlers Your robots.txt Was Never Configured For

Google added Google-CloudVertexBot to its official list of crawlers in 2026 — a new user-agent specifically for sites owners that build Vertex AI Agents and want that agent crawling their own site on their behalf. It is one of a growing list of newer, purpose-specific AI crawlers most robots.txt files have simply never been updated to address at all.

SS
Jul 17, 2026
Get AEO updates weekly.

Citation shifts, algorithm changes, and what's actually working.