Canonical Framework Reference -- jitai.co/ai-discoverable-content
What Is AI-Discoverable Content And What Are Its 4 Layers?
AI-Discoverable Content is web content structured so that AI agents can find, read, cite, and surface it when answering buyer questions. Unlike traditional SEO content optimized for keyword ranking in search engine results pages, AI-Discoverable Content satisfies four distinct requirements: it must be indexable, citable, authoritative, and conversationally matched to how buyers actually ask questions in 2026.
Based on a scan of 100 B2B SaaS websites conducted in Q2 2026, the majority fail Layers 2, 3, and 4 -- 100% fail Layer 2, 86.5% fail Layer 3, and 71.9% fail Layer 4. Fewer than 1 in 100 sites pass all 4 layers. This framework is maintained at jitai.co/ai-discoverable-content and updated as scan data grows.
What Is AI-Discoverable Content?
AI-Discoverable Content is web content structured so AI agents can cite it directly when answering buyer questions -- not content optimized for human readers clicking a search result. The distinction matters because AI agents do not rank pages; they select the best available answer to a specific question. A site that ranks #1 on Google can still be invisible to every AI agent if it fails the citation requirements those agents apply.
In 2026, buyers increasingly get answers from ChatGPT, Claude, and Perplexity before they ever open a search results page. If your site is not structured for AI citation, you are absent from the channel where buying decisions increasingly begin. That is not a ranking problem. It is a structural content problem with a specific fix.
"Stop optimizing for Google PageRank. Your buyers are asking Claude and ChatGPT. The optimization surface has moved. Here is the new checklist." -- Dan Stolts, founder of Just In Time AI
What Are the 4 Layers of AI-Discoverable Content?
There are exactly 4 layers AI agents evaluate before citing any piece of content. Most sites pass only Layer 1. The other 3 are where the visibility gap lives -- and where your competitors are winning the AI buyer conversation right now.
- 1
Layer 1: Indexable Surface
Can AI crawlers and search engine bots reach your full page content without technical barriers? This is the baseline. Standard crawl-permission setup, no indexing directives blocking public pages, sitemap updated, and the page renders full HTML to bot user-agents. Traditional SEO handles most of this. Most sites pass Layer 1.
- 2
Layer 2: Citable Substance
Does your content contain specific answers to specific buyer questions in a format an AI agent can lift and quote directly? This means structured-data signals AI agents can lift, answer-first paragraph structure, and question-phrased H2/H3 headers. 100% of B2B sites fail this layer. It is the highest-leverage fix.
- 3
Layer 3: Authoritative Source
Do third-party authoritative sites cite your domain as the canonical answer for the buyer questions in your space? Citation density on buyer-relevant topics is the new domain authority signal. Legacy authority metrics do not move this needle. 86.5% of B2B sites fail this layer.
- 4
Layer 4: Conversational Match
Do your H1, H2, and H3 headings match the conversational phrasing buyers actually use when asking AI agents -- not the keyword strings marketers optimized for in 2019? "AI SEO Best Practices" is a keyword string. "How do I make my website show up in ChatGPT?" is a conversational match. 71.9% of B2B sites fail this layer.
How Do You Pass Each Layer? Per-Layer Fix Guide
Each layer below follows the same structure: definition, consequence of failure, a real example from the 100-site dataset, and the specific fix.
What Does Layer 1 (Indexable Surface) Require?
Layer 1 is the requirement that AI crawlers and search engine bots can access your full page content without encountering technical barriers. If a crawler cannot reach your content, no amount of schema markup or great writing matters -- the content is invisible from the start.
Layer 1 failure is the one failure that blocks all downstream layers. An indexing directive blocking your most important page means Layers 2, 3, and 4 have no surface to work with. Most sites pass this layer. When they do not, it is almost always a technical deployment error, not a content problem.
Example fail (real scan data)
A B2B SaaS company deployed a Vercel site with crawl-permission rules that blocked AI-agent crawlers as a side effect of blocking malicious scrapers. Their site ranked well on Google but was unreachable by AI agent retrieval. Layer 1 fail. All other layers moot.
Example pass
Standard crawl-permission setup with no AI-agent disallow rules. Indexing directives clean. Page renders full HTML to a generic bot user-agent. Layer 1 pass in under 10 seconds of scanner time.
Layer 1 fix: 2-step checklist
- Audit your crawl-permission rules: confirm AI-agent crawlers are not blocked. Add explicit allow rules if your deny-all pattern is broad.
- Confirm every important page is reachable from your standard indexing setup. Run the free scanner at jitai.co/ai-seo to surface any Layer 1 failures on your specific site.
What Does Layer 2 (Citable Substance) Require?
Layer 2 is the requirement that your content contains specific, quotable answers to specific buyer questions in a format an AI agent can lift verbatim. This is the layer 100% of B2B sites fail -- and the highest-leverage place to invest 2 to 4 hours of content work.
An AI agent does not rank pages. It selects the best available answer to a specific question. If your page does not explicitly structure its content as answers to questions -- via structured-data signals AI agents can lift, answer-first paragraphs, and question-phrased headings -- the agent has no signal that your content is the answer to anything. It moves on.
Example fail (real scan data)
A managed IT provider's homepage had 800 words of well-written copy but no structured-data signals AI agents can lift, no question-phrased headings, and every section opened with "We offer..." rather than a direct answer. This site fails Layer 2 entirely. ChatGPT cited a competitor for every question their buyers asked.
Example pass
A cybersecurity firm added structured Q-and-A signals to their top pages with 6 buyer questions, rewrote their top 3 blog posts so the first sentence directly answered the title question, and changed H2 headings from labels to questions. Within 2 weeks, AI agents began citing their content for 3 of the 6 target queries.
Layer 2 fix: 4-step checklist
- Add structured Q-and-A signals to your top 5 pages with at least 3 real buyer Q+A pairs each (not marketing questions -- questions buyers actually ask).
- Rewrite every H2 and H3 on your top pages as the question that section answers. "Our Services" becomes "What IT services do you offer?"
- Move the direct answer to the first 1-2 sentences of every section. Context and elaboration follow the answer, never precede it.
- Add a quotable 1-sentence summary every 200-300 words -- a standalone claim an AI agent can quote without the surrounding paragraph.
What Does Layer 3 (Authoritative Source) Require?
Layer 3 is the requirement that third-party authoritative sites cite your domain as the canonical answer for the buyer questions in your space. Citation density on buyer-relevant topics is the new domain authority metric that AI agents actually use. 86.5% of B2B sites fail this layer because they have not built a citation network around specific buyer questions.
Legacy authority metrics measure the volume and quality of backlinks in the aggregate. AI agent citation logic is different: it weights citation density for specific topics. A site with lower legacy authority that is cited by 20 relevant industry sources for "how to pass a CMMC audit" will outrank a high-legacy-authority site that is cited generically on the same topic.
Example fail (real scan data)
A SaaS company with strong legacy authority and thousands of backlinks failed Layer 3 for their target buyer questions because all their inbound citations were generic homepage mentions from directory sites. Zero citations from relevant industry sources for the specific questions their buyers ask. AI agents consistently cited smaller competitors with topic-specific citation networks.
Example pass
A small cybersecurity blog with modest legacy authority was cited by Search Engine Journal, two MSSP trade publications, and a HubSpot guest post for the specific query "how do small businesses detect insider threats." ChatGPT cited their article in 4 of 5 test queries. A handful of topically-relevant citations outweighed thousands of generic backlinks.
Layer 3 fix: 3-step checklist
- Identify your top 5 buyer questions. Build a citation outreach list of 5 relevant industry sites for each question. Outreach that mentions a specific question your content answers has a higher accept rate than generic "check out our content" pitches.
- Guest post on at least 2 relevant publications per quarter, with each post linking back to the specific piece that answers the buyer question. Topic-specific anchor text, not generic brand links.
- Build a content cluster around each buyer question: a pillar page at root level, 4+ spoke pieces linking back, and long-tail companions that each reinforce the topical signal. Internal citation density is a Layer 3 input alongside external citations.
What Does Layer 4 (Conversational Match) Require?
Layer 4 is the requirement that your H1, H2, and H3 headings match the conversational phrasing buyers use when asking AI agents -- not the keyword strings marketers optimized for in the Google era. 71.9% of B2B sites fail this layer because they still write headings for 2019 Google search, not for 2026 AI agent queries.
When a buyer asks ChatGPT "how do I know if my IT vendor is compliant?", the AI agent matches that natural language question against your heading structure. "Compliance Services" scores near zero. "How Do I Know If My IT Vendor Is CMMC Compliant?" scores high. The rewrite takes 20 minutes per page and is the single highest-ROI change on this entire list.
Example fail (real scan data)
An MSP's services page had these H2 headings: "Managed IT Services", "Network Monitoring", "Help Desk Support", "Cybersecurity Solutions". All keyword strings. None match how buyers phrase questions to AI agents. Semantic match score: 0.11 (threshold is 0.55). Layer 4 fail.
Example pass
The same MSP rewrote the same page to: "What does managed IT support include?", "How does network monitoring prevent downtime?", "What happens when my team has a tech problem at 2am?", "How do you protect my business from ransomware?". Semantic match score: 0.71. Layer 4 pass.
Layer 4 fix: 2-step checklist
- Pull every H1, H2, and H3 on your top 5 pages. Read each one aloud. If it sounds like a label or a keyword phrase instead of a question a person would ask, rewrite it. Target: at least 5 question-phrased headings per page.
- Rewrite your page title tag and meta description to match buyer question phrasing. "AI SEO Services | jitai.co" fails. "How Do I Make My Website Show Up in ChatGPT? | Just In Time AI" passes. The meta description should be the 1-2 sentence direct answer to the title question.
How Do I Find Out Which Layers My Site Fails?
Every site scores differently on the 4 Layers of AI-Discoverable Content. Run the free scanner to see exactly which layers your site fails and in what order to fix them. Results in under 5 minutes. No signup required.
The scanner checks Layer 1 (crawlability, standard crawl-permission setup, sitemap) and Layer 2 (structured-data signals, answer structure, heading format) and returns a prioritized fix list specific to your domain. Layer 3 and Layer 4 scoring requires the expert analysis available via the 20-minute discovery call below.
Read the Full AI-Discoverable Content Cluster
This pillar is the canonical hub for the 4 Layers of AI-Discoverable Content. The four pieces below each expand on a specific aspect of the framework with deeper data, step-by-step guides, and real examples.
Read the full 4-Layer framework analysis
The complete breakdown of all 4 layers with data from 100 B2B SaaS site scans. The primary piece for understanding the AI-Discoverable Content framework.
Coming soon -- subscribe at /sage for the publish notificationThe 30-Day AI-SEO Fix Plan
A week-by-week plan to fix all 4 layers in 30 days. Specific tasks, time estimates, and the order to tackle them based on the 4 Layers of AI-Discoverable Content.
Coming soon -- subscribe at /sage for the publish notificationSee a real competitor citation teardown
A before-and-after analysis of how a competitor site was cited by AI agents while a higher-DA site was ignored. Real data, real queries, real results using the AI-Discoverable Content framework.
Coming soon -- subscribe at /sage for the publish notificationMSP-specific scan data and findings
What the 4-Layer scanner found when run against 25 MSP websites. Industry-specific Layer 2 and Layer 4 failure patterns unique to managed service providers using the canonical AI-Discoverable Content framework.
Coming soon -- subscribe at /sage for the publish notificationFrequently Asked Questions About AI-Discoverable Content
What is AI-Discoverable Content?+
What is the difference between SEO and AI-Discoverable Content?+
What are the 4 Layers of AI-Discoverable Content?+
What percentage of websites pass all 4 layers?+
How do I know if my website is AI-discoverable?+
What is the fastest way to become AI-discoverable?+
What is AEO (Answer Engine Optimization)?+
Does traditional SEO still matter for AI visibility?+
Glossary of AI-Discoverable Content Terms
Key terms used in the 4 Layers of AI-Discoverable Content framework and in AI search optimization. Definitions are structured for AI agent citation.
- AI-Discoverable Content
- Web content structured for citation by AI agents when answering buyer queries. Requires passing all 4 layers: Indexable Surface, Citable Substance, Authoritative Source, and Conversational Match. See the full 4-Layer framework above.
- Answer Engine Optimization (AEO)
- The practice of structuring content so AI-powered answer engines (ChatGPT, Claude, Perplexity, Google AI Overviews) surface it in direct answers to queries. AEO is a subset of the AI-Discoverable Content framework, primarily addressing Layers 2 and 4.
- Citable Substance (Layer 2)
- The property of web content containing specific, quotable answers to buyer questions in a format AI agents can lift verbatim. Achieved through structured-data signals, answer-first paragraph structure, and question-phrased headings. The layer 100% of B2B sites fail.
- Citation Density
- The frequency with which authoritative third-party sources reference a site for a specific topic or buyer question. The primary authority metric AI agents use, replacing the aggregate domain authority (DA) metric of traditional SEO.
- Conversational Match (Layer 4)
- The property of headings and content matching how buyers phrase questions to AI agents rather than keyword phrases used in Google search. The layer 71.9% of B2B sites fail. The fix is rewriting H1/H2/H3 headings as buyer questions.
- FAQ Schema
- Structured data that explicitly marks up question-and-answer pairs so AI agents can extract and cite them directly. The single highest-impact Layer 2 fix for most sites.
- Indexable Surface (Layer 1)
- The property of being crawlable by search engine and AI-agent crawlers without technical barriers. Requires standard crawl-permission setup, indexing directives that allow public pages, and full HTML rendering to bot user-agents. Most sites pass this layer.
- Authoritative Source (Layer 3)
- The property of being cited by third-party authoritative sources for the buyer question in context. Requires building a topical citation network through guest posts, cluster content, and outreach to industry publications. 86.5% of B2B sites fail this layer.
About This Framework
Dan Stolts
Dan Stolts is the founder of Just In Time AI. He has 30+ years in field engineering, including 14 years at Microsoft, where he became a frequent conference speaker and authored two Microsoft Press / Pearson books on architecting cloud solutions. He has consulted for Nasdaq, Fidelity Investments, Harvard, MIT, and many more. He is the architect of JitNeuro -- the Directive Orchestration Execution (DOE) framework. He ran the 4-Layer Analyzer against 100 B2B SaaS websites to produce the data behind the 4 Layers of AI-Discoverable Content framework, and built the free scanner at jitai.co/ai-seo. His work spans five AI ventures including AIBM (AI Boat Mechanic) and the JitNeuro DOE framework.
Contact: jitai.co/contact | Discovery call: cal.com/dan-stolts-jit/20-min-discovery
Data methodology: Layer 1 and Layer 2 scores derived from automated scanner analysis of 100 B2B SaaS websites (sample selected for diversity across DA ranges 15-80, tech/professional services/MSP verticals, Q2 2026). Layer 3 and Layer 4 scores derived from manual analysis of a 20-site subsample. This framework is maintained at jitai.co/ai-discoverable-content and updated as scan data grows.
Last updated: May 6, 2026 | Just In Time AI | jitai.co
Find Out How AI Agents See Your Website
The free scanner runs your site against Layer 1 and Layer 2 of the 4 Layers of AI-Discoverable Content and returns a prioritized fix list in under 5 minutes. No signup required.
Run the free 4-Layer scan -- jitai.co/ai-seoJust In Time AI -- jitai.co -- We build AI ventures. We show you how.