Back to Knowledge Hub
戦略・ポジショニングブランドポジショニングフレームワーク初級

The AI Librarian's Recommendation List — AEO, GEO, and LLMO Are Different Librarians in the Same Library

AEO, GEO, LLMO, and MEO are different terms, but they reference the same data. They're simply different AI librarians recommending books from the same Google-organized library. This unified framework simplifies AI-era marketing strategy.

AEOGEOLLMOAI CognitionUnified Framework

What You'll Learn

You will resolve the confusion around AEO, GEO, LLMO, and related terminology, and understand AI-era marketing through a single unified framework. This article provides the Library & Librarian metaphor you can use directly when explaining to clients or colleagues.

Prerequisites

This article is beginner-friendly. For detailed definitions of each term, see the AEO Glossary, GEO Glossary, and LLMO Glossary. For the foundational concept of AI Cognition, we recommend reading What is AI Cognition? first.

Main Content

The Terminology Confusion

Terms related to AI search optimization are proliferating. AEO (Answer Engine Optimization), GEO (Generative Engine Optimization), LLMO (Large Language Model Optimization), and MEO (Map Engine Optimization). Each was coined by different companies and researchers with different definitions.

However, these terms are fundamentally describing the same phenomenon from different angles.

The Unified Library & Librarian Model

Let's resolve this confusion with a single metaphor.

Google is the person who organized the world's largest library.

Over 20+ years, Google crawled, indexed, labeled, and shelved websites from across the internet. PageRank measured each book's credibility. Structured data provided machine-readable labels. E-E-A-T evaluated author expertise. This massive catalog system is Google's search engine.

AI platforms are the librarians.

ChatGPT, Perplexity, Google Gemini, and Bing Copilot are librarians who consult this library's collection to answer visitor questions. Each librarian has their own preferences — one favors academic sources, another prefers practical guides. But they're all referencing the same library.

Mapping each term to this unified model:

TermMeaning in the Unified ModelFocus
SEOOptimizing labels in the libraryPosition on Google's shelves
AEOGetting the librarian to recommend youRecommendation in AI answers
GEOBecoming a book the librarian citesCitation in AI references
LLMOBecoming a book the librarian remembersRecognition in AI internal models
MEOGetting recommended at the local branchDisplay in map-based search

Same data. Different access methods.

Why This Unified Understanding Matters

If you think each term requires separate strategies, your efforts will fragment. Creating separate "AEO teams," "GEO teams," and "SEO teams" is inefficient.

Based on the unified model, what you need to do converges to two things:

1. Secure your position on the library shelves (SEO foundation)

No matter how good your book is, the librarian can't recommend it if it's not on the shelves. Being indexed by Google and achieving a minimum search ranking is the prerequisite.

Real data supports this: pages ranking below position 60 on Google are almost never cited by AI search engines. SEO is a necessary condition for AI Cognition.

2. Create books the librarian wants to recommend (content quality)

Being on the shelf isn't enough. The librarian needs to think "this book is perfect for this visitor's question." Specifically:

  • Structure that directly answers questions: Titles in "What is X?" or "How to do Y" format
  • Credible data backing: Statistics with source attribution, original research
  • Self-contained answer blocks: 134-167 word paragraphs that completely answer one question
  • Third-party perspective: "The industry shows..." not "Our company is great..."

Librarian Preferences Vary — Platform Differences

Within the unified model, each librarian has their own "taste":

Librarian (AI)Preferred Source TendenciesCharacteristics
ChatGPTTraining data + web searchHeavily influenced by training data. Retains older information
PerplexityReal-time web searchStrong on current information. Shows source URLs explicitly
Google GeminiGoogle search results + Knowledge GraphStrongly dependent on Google search rankings
Google AI OverviewsTop Google results + structured dataMost influenced by structured data

But the commonalities outweigh the differences. No librarian can recommend a book that's not in the library. Every librarian prioritizes trustworthy sources. Every librarian prefers content that matches the question format.

Therefore, an approach where 80% of tactics are shared and 20% are platform-specific adjustments is most efficient.

Why SEO People Should Care About AI Cognition

"If I'm doing SEO, isn't AI Cognition covered?" is a common question. The answer is half right, half wrong.

SEO is the prerequisite. If your book isn't on the library shelves, no librarian can recommend it. This part is correct.

But SEO alone isn't sufficient. According to Ahrefs research, only 12% of AI-cited URLs rank in Google's top 10. This means even a #1 Google ranking doesn't guarantee AI citation.

Key differences between SEO and AI Cognition:

AspectSEOAI Cognition
GoalTop rankings in search resultsRecommendation and citation in AI answers
MetricsSearch rank, CTR, organic trafficMention rate, Citation rate, recommendation context alignment
Content designKeyword-centricQuestion-answer pair structure
LabelingTitle tags, meta descriptionsStructured data, entity recognition
Measurement difficultyLow (rankings are numbers)High (designing what to measure is itself the challenge)

Unified Framework: 3-Step Execution

A framework for improving AI Cognition in one flow:

Step 1: Secure your shelf position (SEO foundation)

  • Submit sitemap to Google Search Console
  • Clear basic technical SEO requirements (checklist)
  • Achieve at least top 30 ranking for key terms
  • Acquire external backlinks

Step 2: Write books the librarian wants to recommend (content structuring)

  • Use "What is X?" and "How to Y" title formats
  • Include self-contained answer blocks in each section
  • Add source attribution to all statistics
  • Implement structured data
  • Strengthen E-E-A-T signals

Step 3: Measure the librarian's recommendations (AI Cognition monitoring)

  • Regularly query AI platforms with your relevant questions and check recommendation status
  • Measure both branded and industry query Mention and Citation rates
  • Verify recommendation context aligns with your positioning
  • See AEO Monitoring Tools Guide for details

Frequently Asked Questions

Can small companies get recommended by the AI librarian?

Yes. The AI librarian decides recommendations based on "how well the book fits the visitor's question," not the publisher's size. Small, specialized sites with 40-50 educational articles have been confirmed to be cited by AI over major media outlets.

Should I do separate AEO and GEO strategies?

No. The same tactics cover both. "Getting on the shelves (SEO)" and "creating books the librarian wants to recommend (content quality)" are shared. Platform-specific adjustments (detailed structured data implementation, etc.) account for only about 20% of the work.

Action Checklist

  • Confirmed your site is indexed by Google
  • Submitted sitemap to Google Search Console
  • Checked search rankings for key industry terms (target: within top 30)
  • Verified article titles use "What is X?" or "How to Y" format
  • Confirmed key articles have self-contained answer blocks (134-167 words)
  • Checked recommendation status on 3+ AI platforms (ChatGPT, Perplexity, Gemini)

Next Steps