SEO remains the foundation
AI visibility starts with crawlable, indexable, technically sound pages. AEO and GEO extend SEO; they do not replace it.
public knowledge base · SEO / AEO / GEO
The field guide for AI-era search visibility.
This is the public operating model behind LLMRanks: technical SEO, answer engine optimization, generative engine optimization, AI citations, brand mentions, source attribution, and the evidence we use to keep recommendations current.
01 · current operating model
What we believe today, until the evidence changes.
The top of this page should change as the market changes. Older findings stay below as an evidence trail, with corrections and confidence shifts rather than silent rewrites.
Fan-out is the unit of coverage
A page should answer the cluster of sub-questions engines generate around a query, not just repeat one head keyword.
Consensus shapes recommendations
Reviews, Reddit, YouTube, editorial lists, partner pages, and brand mentions influence what AI says about a company.
Extractable evidence wins citations
Statistics, comparisons, definitions, code examples, primary research, and clear source attribution make content easier to quote.
01 · Access layerTechnical SEO foundations12k tokens
Jul 3, 2026+
Crawlability, indexability, canonicals, robots directives, sitemaps, redirects, rendering, and AI crawler access rules.
what this helps decide
- Can search engines and AI crawlers reach the canonical document?
- Which pages should be indexed, deindexed, redirected, or consolidated?
- Where are crawl traps, duplicate URLs, and render-blocking choices hiding?
section inventory
Technical SEO Foundations 2026: An Engineering Reference1. Crawlability2. Indexability3. Sitemaps4. URL Structure5. HTTPS + Security Headers6. Redirect Strategy
source
opus-deepdive-2026-05-09
approx size
46,220 chars
02 · Meaning layerOn-page SEO7.7k tokens
Jul 3, 2026+
Titles, headings, first answers, semantic HTML, internal anchors, entity clarity, media text, and snippet control.
what this helps decide
- Does the page say the answer in the language users and engines retrieve?
- Are headings and intros organized for scanning, extraction, and trust?
- Does the page expose enough text around images, videos, and key entities?
section inventory
Title Tags in 2026Meta DescriptionsHeading Hierarchy (H1–H6)Internal Anchor TextImage OptimizationContent StructureOn-Page User Signals
source
opus-deepdive-2026-05-09
approx size
30,680 chars
03 · Entity layerSchema.org / Structured Data12k tokens
Jul 3, 2026+
Schema.org, JSON-LD, rich-result eligibility, entity disambiguation, Organization/Product/Article data, and schema myths.
what this helps decide
- Which schema creates real rich-result eligibility?
- Where does JSON-LD clarify entities without pretending to be an AI citation lever?
- Are author, organization, product, review, FAQ, and breadcrumb entities consistent?
section inventory
Schema.org Structured Data: The Definitive 2026 Reference for SEO + AEOFormat Wars: JSON-LD vs Microdata vs RDFa (Settled)The @id Pattern (Most Audit Tools Miss This)Per-Type Deep DiveValidation Tooling (2026 State)What Actually Drives Google Rich Results vs LLM Citations (2026 Scorecard)AEO-Critical Markup Patterns by Content Type
source
opus-deepdive-2026-05-09
approx size
49,868 chars
04 · Authority flowInternal linking strategy7.0k tokens
Jul 3, 2026+
Contextual links, hub architecture, orphan detection, cannibalization, anchor text, and PageRank flow through large sites.
what this helps decide
- Which pages deserve more contextual in-body links?
- Where do hubs, glossary pages, and comparison pages route topical authority?
- Which pages compete with each other instead of reinforcing a topic?
section inventory
Comprehensive Internal Linking Strategy in 2026Site Architecture: Hub-and-Spoke vs Silo vs FlatLink Equity Flow & PageRank DistributionAnchor Text StrategyLink Attributes: dofollow / nofollow / sponsored / ugcInternal Link Count Per PageLink Placement & Equity Weight
source
opus-deepdive-2026-05-09
approx size
28,044 chars
05 · Evidence layerContent strategy + E-E-A-T7.5k tokens
Jul 3, 2026+
Original research, expertise signals, content depth, factual density, first-hand examples, and anti-scaled-content safeguards.
what this helps decide
- Does the content add information that did not already exist everywhere else?
- Are claims specific, attributed, dated, and useful enough to be quoted?
- Where should AI assist drafting without creating scaled-content risk?
section inventory
Content Strategy for SEO + AI Search in 2026E-E-A-T: Concrete Demonstration, Not Abstract ClaimsHelpful Content + People-First (Post-March 2024 Integration)Topical Authority + Content ClustersContent Depth: Word Count vs. DepthContent FreshnessOriginality + Duplicate Content
source
opus-deepdive-2026-05-09
approx size
30,040 chars
07 · Experience layerCore Web Vitals + Page Experience7.6k tokens
Jul 3, 2026+
INP, LCP, CLS, performance budgets, rendering tradeoffs, image loading, and page-experience diagnostics.
what this helps decide
- Are pages fast and stable enough for users and crawlers?
- Which template or asset choices degrade retrieval, rendering, or conversion?
- How should performance fixes be prioritized by page type and revenue impact?
section inventory
Page Experience + Core Web Vitals in 2026: The Engineering ReferenceLCP — Largest Contentful PaintINP — Interaction to Next PaintCLS — Cumulative Layout ShiftLab Metrics — TTFB, FCP, TBTMobile vs Desktop ScoringHosting + CDN Choices (2026 Landscape)
source
opus-deepdive-2026-05-09
approx size
30,544 chars
08 · Citation layerAEO / GEO — Answer Engine + Generative Engine Optimization7.6k tokens
Jul 3, 2026+
Answer engine optimization, generative engine optimization, fan-out coverage, answer-first passages, and citation readiness.
what this helps decide
- Which fan-out sub-questions must this page answer directly?
- What makes a passage easy to cite without reducing quality for humans?
- Which claims need statistics, comparisons, code, quotes, or original data?
section inventory
How LLMs Actually Select Content to Cite in 2026What Makes Content Highly CitableStructural Patterns LLMs FavorKey TakeawaysContent Depth and Length Sweet SpotsInternal vs External Citation Behavior by EngineBrand Mentions, Entity Recognition, and the Knowledge Graph
source
opus-deepdive-2026-05-09
approx size
30,380 chars
09 · Engine layerAI search engines individually8.8k tokens
Jul 3, 2026+
ChatGPT, Gemini, Claude, Perplexity, Grok, Google AI Overviews, AI Mode, retrieval behavior, and source differences.
what this helps decide
- Which engines reward owned pages, third-party pages, forums, video, or reviews?
- How should measurement differ when engines cite, name, summarize, or omit links?
- Where do source mixes change fast enough to require monitoring instead of dogma?
section inventory
AI Search Engine Retrieval & Citation Mechanics: 2026 Field GuideChatGPT (OpenAI Search + Browsing)Claude (Anthropic) with Web SearchPerplexityGoogle Gemini + AI Overviews + AI ModeMicrosoft Copilot / Bing ChatCross-Platform Synthesis for Engineering Teams
source
opus-deepdive-2026-05-09
approx size
35,320 chars
10 · Future-proofing2026 modernization + future-proofing7.9k tokens
Jul 3, 2026+
Modern 2026 guidance, AI search product changes, llms.txt posture, snippet eligibility, policy shifts, and practices to retire.
what this helps decide
- Which older SEO recommendations are now obsolete or risky?
- What is cheap to support, but weak enough that it should not be sold as magic?
- Where should recommendations change as Google and AI engines update products?
section inventory
What's New in SEO + LLM Optimization for 2026Site NameBlock low-value training crawlersExplicit allows for retrieval (some bots only crawl if explicitly named)
source
opus-deepdive-2026-05-09
approx size
31,648 chars
11 · Agent layerAgentic experiences — designing for AI agents that operate your site4.7k tokens
Jul 3, 2026+
Designing sites for AI agents that compare, click, fill forms, read policies, and complete tasks on behalf of users.
what this helps decide
- Can an agent understand offerings, constraints, pricing, and next actions?
- Where should structured flows supplement ordinary landing pages?
- What site affordances reduce ambiguity for both users and automation?
section inventory
Agentic Experiences — Designing for AI Agents That *Operate* Your SiteThe Shift: From Crawlers Reading to Agents DrivingHow Agents See a Page (vs How a Crawler Sees It)What Determines Whether an Agent Can Use Your SiteWhat Breaks Agents (Anti-patterns)Detection — How to Know if Agents Are Reaching YouWhat This Means for Audit + Strategy
source
opus-deepdive-2026-05-09
approx size
18,740 chars
04 · evidence log
The lower page grows like a public lab notebook.
The top operating model should stay clean. The trail below can keep expanding with studies, teardowns, verification passes, and videos that explain what changed.
Supabasestudy finding · Jun 22, 2026
Citation Study — 270 AI Answers (June 2026)
Stored as `citation-study-2026-06` with 613 tokens and tags aeo, citations, study.
PrimaryPosition teardown
A competitor analysis found a glossary-plus-sitewide-linking model ranking many long-tail SEO terms, but also exposed a gap in contextual in-body links.
270-answer citation study
Cross-engine overlap was tiny, Reddit led citations, YouTube concentrated in AI Overviews how-to answers, and named-brand visibility diverged from source linking.
AEO/GEO verification pass
The KB downgraded blanket FAQ-ification, softened schema-as-AI-lever claims, strengthened fan-out coverage, and elevated original research.
Practitioner channel digest
A synthesis of recent AI SEO commentary captured market disagreement around prompt tracking, schema, llms.txt, review influence, and off-site mentions.
Publish the model
Explain the recommendation logic plainly, including where confidence is high or still contested.
Keep the history
When guidance changes, append the update and adjust the top summary instead of burying the old claim.
Separate proof from practice
Label primary research, platform documentation, practitioner inference, and internal experiments differently.
Turn updates into media
Every meaningful change can become a short video, a YouTube explainer, and a shareable visual.
05 · source library
External references behind the public guidance.
This list should keep expanding. The rule: public claims get public receipts wherever possible.
Google Search Central: AI features and your websiteEligibility, controls, AI Overviews, AI Mode, and fan-out mechanics.
Google Search Central: optimizing for generative AIGoogle's current guidance: no special machine-readable files or special schema required.
Google Search Central: helpful, reliable, people-first contentThe editorial quality baseline for search and AI-search visibility.
Ahrefs: AI Overview citations vs organic top 10Rank/citation decoupling and citation overlap analysis.
Ahrefs: brand visibility correlations in AI OverviewsBrand mentions, backlinks, and other correlation signals.
Vercel: the rise of the AI crawlerCrawler behavior, request volume, and JavaScript rendering caveats.
Semrush: most-cited domains in AILarge-scale citation volatility and domain-level source patterns.
Otterly: llms.txt experimentServer-log evidence on llms.txt crawler usage.
next step