Hidden RadiusSearch intelligence beneath the visible result
Home / Modern Search Comparison Matrix | Hidden Radius
Purpose, ownership and limits

Modern Search Comparison Matrix

Use this matrix to classify a proposal, page, metric or technical decision before accepting the terminology around it.

Published 2026-07-22Last reviewed 2026-07-22Primary-source analysis
The matrix

Nine terms across the discovery stack

TermCategoryPrimary jobApplied byMain inputExpected outcomeDoes not guarantee
SEOPublishing/search disciplineCrawl, index, understand and rank useful pagesPublisher, developer, marketerPublic pages, architecture, links, contentSearch visibilityNo guaranteed ranking
GEOGenerated-answer strategyImprove source usefulness or visibility in generated answersPublisher, organization, marketerOriginal evidence, clarity, sources, entity contextGenerated-answer inclusion or citationNo universal cross-platform formula
AEOAnswer-focused publishingSatisfy a specific question directlyPublisher, support team, educatorQuestion headings, immediate answers, conditionsAnswerabilityNo guaranteed selection
AIOAmbiguous umbrella/product labelAI Optimization or Google AI OverviewsDepends on definitionScope-specificScope-specificNo settled meaning without context
REGEntity and information guidance frameworkRecognize the correct entity and retrieve correct factsBusiness, publisher, data stewardNames, relationships, facts, official linksRecognition and retrievabilityNot a platform standard or guarantee
RAGTechnical AI architectureRetrieve outside knowledge during generationAI engineer, product teamIndex, retriever, context, generatorGrounded generationBad sources and retrieval still fail
LLMLanguage modelInterpret and generate languageModel and product developerTraining, prompt, context, toolsGenerated outputNot a complete live search system
Structured DataMachine-readable vocabularyLabel page meaning and relationshipsPublisher, developerJSON-LD aligned with visible contentMachine interpretation and feature eligibilityDoes not prove claims or ranking
robots.txtCrawler instruction fileCommunicate access rules to compliant crawlersSite ownerUser-agent, allow and disallow rulesCrawl permissionNot authentication, indexing or ranking
How to use it

Three questions before approving work

  1. What category is this? A publishing discipline, technical architecture, platform feature, protocol or original framework should not be measured the same way.
  2. Who controls the main input? A site owner can change content and access; only a platform operator can change its retrieval and ranking system.
  3. What outcome is actually observable? Indexing, citation, accurate representation and customer action are different results.
Procurement warning

Acronyms can hide undefined scope

A proposal that promises “AIO” or “GEO ranking” without defining the platforms, inputs, monitoring method and limits is not sufficiently scoped. Hidden Radius recommends rewriting the work in ordinary operational language before evaluating cost or performance.

Downloadable data

Use the matrix internally

The complete comparison is also included in the package as a CSV file for training, planning and audit worksheets. The dataset preserves the category, owner, inputs, outcomes and limitations expressed on this page.

Source trail

Primary material behind this analysis

Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks

The original RAG paper combining parametric generation with retrieved non-parametric memory.

Open the original source at arXiv →

Related reading

Continue through the discovery stack

Hidden Radius Knowledge Centre

Navigate the complete Hidden Radius knowledge centre covering modern search, AI discovery, retrieval, entities, standards, platforms, research and implementation.