100 guides
AI Search
AI answer engines do not behave like search engines, and the differences are measurable. Across 90 AEOeye audits the same brand was named by Claude in 31% of buyer questions but by Google AI Overviews in only 2% — checking one engine tells you very little.
How to read this section
- 1How each engine sources its answer: live retrieval, training memory, or a blend.
- 2Why the same question returns different brands on different engines.
- 3What causes an answer to change run-to-run, and how to tell drift from a real improvement.
AI Search
AI Search Metrics SQL Starter Kit: Compute Mentions and Citations Locally
Run transparent SQLite queries against AI-search audit exports to separate answer availability, brand mentions, recommendations, and citation denominators.
Sep 22, 2026 · 7 min read
AI Search
AI Recommendation Stability Calculator: Measure Repeat-Run Agreement
Use a browser-only calculator to compare repeated AI brand recommendations with exact-set agreement, pairwise Jaccard, coverage, and a deterministic bootstrap interval.
Sep 22, 2026 · 8 min read
AI Search
AI Citation Parser Test Corpus: 24 Synthetic Edge Cases With Expected Labels
A small, inspectable JSONL corpus for testing AI citation extractors across numbered markers, footnotes, URLs, malformed syntax, and ambiguous claim scope.
Sep 22, 2026 · 8 min read
AI Search
AI Brand Co-Mention Network Generator: Build a Competitor Edge List
Turn coded AI recommendation lists into a transparent, weighted co-mention network with a tested dependency-free Python utility and synthetic CSV fixture.
Sep 22, 2026 · 9 min read
AI Search
AI Source Ownership Register: Map Syndication Before Counting Diversity
A practical register for tracing AI-search citations through URLs, publishers, owners, canonical sources, and syndication relationships before claiming source diversity.
Sep 16, 2026 · 8 min read
AI Search
AI Search Study Dataset Card: A Template for Auditable Results
A copyable dataset-card template for documenting AI search studies, from sampling frame and prompts to provenance, licensing, bias, and intended use.
Sep 16, 2026 · 10 min read
AI Search
AI Search Reproduction Package: A File-by-File Manifest
A citation-first manifest for packaging AI search prompts, answers, source captures, annotations, code, checksums, and release metadata without overstating reproducibility.
Sep 16, 2026 · 10 min read
AI Search
AI Search Prompt Paraphrases: A Protocol for Equivalent Test Sets
A citation-first worksheet for building equivalent AI-search prompt variants without silently changing buyer intent, scope, constraints, or answer format.
Sep 16, 2026 · 9 min read
AI Search
AI Search Context Leakage: A Multi-Turn Test Protocol
A citation-first protocol for separating intended conversation context from uncontrolled carryover in AI search tests, with clean-session and primed-session matrices.
Sep 16, 2026 · 9 min read
AI Search
AI Search Audit Versioning: A Change Log That Preserves Comparability
A practical version ledger for documenting AI search audit changes, preserving provenance, and deciding when historical comparisons remain valid.
Sep 16, 2026 · 9 min read
AI Search
AI Citation Redirect Chains: An Audit Protocol for Moving Evidence
A hop-by-hop worksheet for testing whether an AI-cited URL still identifies the same evidence after redirects, canonical hints, and ownership changes.
Sep 16, 2026 · 10 min read
AI Search
AI Citation Marker Codebook: Scope, Position, and Presentation
A citation-first codebook for recording where AI answer markers appear, what text they scope, and whether the visible presentation is inspectable.
Sep 16, 2026 · 10 min read
AI Search
AI Brand Recommendation Eligibility: A Predeclared Register
A citation-first register for defining which brands belong in an AI recommendation audit before collection, analysis, and reporting.
Sep 16, 2026 · 10 min read
AI Search
AI Source Update Propagation Lag: A Repeatable Measurement Protocol
Measure the time from a verified source change to first detection, first citation, and first correct AI answer without pretending an unknown timestamp is exact.
Sep 13, 2026 · 9 min read
AI Search
AI Search Evidence Grading: A Claim-Level Scale for Research and Audits
A claim-level evidence scale for AI search research, with separate axes for claim fit, time relevance, method transparency, provenance, and conflicts of interest.
Sep 13, 2026 · 9 min read
AI Search
AI Search Personalization Confounders: A 20-Field Testing Checklist
A reproducible 20-field capture checklist for testing AI-search results across accounts, memory, location, devices, models, modes, connectors, and experiment order.
Sep 13, 2026 · 8 min read
AI Search
AI Search Missing-Data Codebook: 12 Statuses That Protect Your Denominator
A deterministic 12-code codebook for AI search audits that separates missing observations from negative findings and keeps every denominator explainable.
Sep 13, 2026 · 9 min read
AI Search
AI Search Engine Agreement Metrics: Compare Lists Without Hiding Disagreement
A practical metric selection guide for comparing AI search outputs by set overlap, top-k overlap, rank order, and per-brand coverage without confusing agreement with truth.
Sep 13, 2026 · 10 min read
Other sections
See where you actually stand
Run a free audit and find out whether AI names you — or a competitor — when buyers ask.