SEO & AI Engine Optimization Framework · May 2026

Multi-Engine Trade-offs: when Google and AI engine optimization diverge

A comprehensive installation and audit reference for deciding how to split optimization effort across the major AI citation surfaces. The companion frameworks in this library cover each engine in…

The Cross Engine Optimization Decision Matrix for AI Search Citation: When to Optimize for One Engine at the Cost of Another, and How to Choose

A comprehensive installation and audit reference for deciding how to split optimization effort across the major AI citation surfaces. The companion frameworks in this library cover each engine in depth: Google AI Overviews and AI Mode, OpenAI's ChatGPT Search family, Perplexity's Search and Pages and Spaces and Comet, and the broader AI citation surface spanning Claude, Bing Copilot, Gemini, Grok, and Meta AI. What this framework adds is the cross engine tradeoff matrix. The studies establishing 11 percent inter engine source overlap, 13.7 percent Google AI Overview to AI Mode overlap, 89 percent platform specific citation patterns per the Yext 17.2 million citation analysis, and the divergent source preferences per engine make a single unified AI strategy impossible for most businesses. This document specifies which signals lift which engines, where engines genuinely diverge, the decision tree for priority engine selection, the citation overlap data, when unified strategy works and when it fails, and the tradeoff documentation per engine pair. Dual purpose: installation manual and audit document.

Cross stack implementation note: code samples in this framework are written in plain HTML for clarity. For React, Vue, Svelte, Next.js, Nuxt, SvelteKit, Astro, Hugo, 11ty, Remix, WordPress, Shopify, and Webflow equivalents see framework-cross-stack-implementation.md. For pure client rendered SPAs (no SSR/SSG) see framework-react.md. For Tailwind specific concerns (purge, dynamic classes, dark mode CLS, focus accessibility) see framework-tailwind.md.

1. Document Purpose

1.1 What This Document Is

This is the canonical operational reference for cross engine optimization decisions. The four engine specific frameworks (framework-aioverviews.md for Google AI Overviews and AI Mode, framework-searchgpt.md for the OpenAI surface family, framework-perplexityspaces.md for Perplexity, and framework-aicitations.md for the broader citation surface) specify what to do per engine. This framework specifies which engine to prioritize when budget, content production capacity, schema rework, freshness cadence, brand voice, and audience reach decisions force a choice.

The data forcing the choice: the Yext Q4 2025 analysis of 17.2 million citations across four engines found only 11 percent of cited domains appear across multiple engines, 89 percent are platform specific. The Ahrefs February 2026 study of 730,000 response pairs found 13.7 percent citation overlap between Google AI Overviews and AI Mode despite the two surfaces sharing a Gemini 3 Pro engine. The Profound August 2024 to June 2025 cross platform analysis found only 11 percent of domains cited by both ChatGPT and Perplexity. The BrightEdge weekly AI search insights series across 2025 to 2026 found citation share concentration in the top 10 most cited domains ranging from 18.5 percent for ChatGPT to 26.7 percent for Perplexity, an indicator that the engines source the open web at structurally different breadths.

A single AI strategy that ignores these divergence patterns produces citation lift on one or two engines and silent invisibility on others. The work of this framework is to make the tradeoff explicit, document the choice with a rubric, and revisit the choice quarterly as engine behavior evolves.

1.2 Three Operating Modes

Mode A, Install Mode. Build a multi engine optimization plan with explicit engine priority order and tradeoff documentation. Follow Sections 2 through 14 in order.

Mode B, Audit Mode. Evaluate an existing engagement's current multi engine posture and identify where unrecognized tradeoffs are silently degrading citation share. Skip to Section 13.

Mode C, Hybrid Mode. Audit first, then re sequence the engine specific work per Sections 7 through 11.

1.3 How Claude Code CLI Should Consume This Document

  1. Read Section 2 and collect client variables, especially current per engine citation state and ICP geographic split.
  2. Run the Section 7 priority decision tree to produce the engine priority order for the engagement.
  3. Apply Section 9 unified strategy patterns to capture the universally additive signals.
  4. Apply Section 10 divergent strategy patterns selectively per priority engine.
  5. Use Section 11 to filter the engine choice through the business type lens.
  6. Run Section 12 against the engagement to surface common multi engine mistakes.
  7. Establish Section 13.4 first 90 days tracking before measuring lift.
  8. Apply Section 14 maintenance cadence and generate the Section 14 report.

1.4 Conflict Resolution Rules

Conflict Rule
Client wants citation on all engines without acknowledging tradeoffs Apply Section 7 decision tree. Document the priority order. Set realistic per engine expectations with the data from Section 6.
Tactic that lifts ChatGPT depresses Perplexity citation Acknowledge per Section 10. Choose based on Section 7 engine priority order, not signal preference.
Engine A budget consumes 80 percent of capacity, engine B citation rate drops Healthy if engine A is the priority. Section 8 documents the per engine cost.
Unified strategy produces zero engine specific lift Diagnostic. Re run Section 9 audit to verify unified work is genuinely deployed before adding divergent work.
Client insists on equal effort across engines Caution. Per Section 6, equal effort produces equal mediocrity. Section 11 routes by business type to pick the engine that actually drives the audience.
Brand voice formality required for B2B citation conflicts with conversational tone for ChatGPT Section 10.2. Choose based on Section 11 business type or maintain two content modes per page surface.

1.5 Required Tools

1.6 Relationship to Neighboring Frameworks

This framework is the decision layer sitting between the four engine specific frameworks. It does not duplicate the per engine signals; it sequences them. The substrate doctrine in framework-contentfirst.md is universally required and gates every engine. The four pillars architecture (SEO, AEO, AIO, GEO) lives in SEO-Search-Appearance.md. The feature level CTR economics across the 30 SERP features live in SERP-Optimization.md. Brand voice tradeoffs are in framework-brandvoice.md. Content production capacity tradeoffs are in framework-ai-content-workflow.md. YMYL gating overlays every engine and lives in framework-ymyl.md.

2. Client Variables Intake

# MULTI ENGINE TRADEOFFS FRAMEWORK CLIENT VARIABLES

# --- Business and Site Identity (REQUIRED) ---
business_name: ""
primary_domain: ""
business_industry: ""
business_type: ""                            # "local_service" | "b2b_saas" | "ecommerce" | "content_publisher" | "ymyl_healthcare" | "ymyl_finance" | "ymyl_legal" | "professional_service" | "media_brand"
ymyl_classification: ""                      # "full_ymyl" | "partial_ymyl" | "lite_ymyl" | "non_ymyl"

# --- ICP Geographic Split (REQUIRED) ---
icp_geo_primary: ""                          # "us_only" | "us_global" | "global_no_us" | "regional_us" | "single_metro"
icp_geo_us_share: 0                          # percent of revenue or pipeline from US
icp_geo_eu_share: 0
icp_geo_apac_share: 0
icp_geo_latam_share: 0
icp_geo_other_share: 0

# --- Priority Engine Selection (REQUIRED, filled after Section 7) ---
priority_engine_1: ""                        # "google_ai_overview" | "google_ai_mode" | "chatgpt_search" | "perplexity" | "claude_web" | "bing_copilot" | "meta_ai" | "grok"
priority_engine_2: ""
priority_engine_3: ""
secondary_engines: []                        # engines monitored but not actively optimized for

# --- Current Citation State Per Engine (REQUIRED) ---
google_ai_overview_citation_status: ""       # "regularly_cited" | "occasionally_cited" | "rarely_cited" | "never_cited" | "unknown"
google_ai_mode_citation_status: ""
chatgpt_search_citation_status: ""
perplexity_citation_status: ""
claude_web_citation_status: ""
bing_copilot_citation_status: ""
meta_ai_citation_status: ""
grok_citation_status: ""

# --- Current Citation State Per Engine, Numeric (REQUIRED) ---
priority_query_count: 0                      # 10 to 30 queries the site targets for citation
google_ai_overview_citation_count: 0         # of priority queries, how many cite the site
google_ai_mode_citation_count: 0
chatgpt_search_citation_count: 0
perplexity_citation_count: 0
claude_web_citation_count: 0
bing_copilot_citation_count: 0

# --- Engine Pair Overlap Baseline (RECOMMENDED) ---
shared_queries_aio_chatgpt: 0                # of priority queries cited in both
shared_queries_aio_perplexity: 0
shared_queries_chatgpt_perplexity: 0
shared_queries_aio_ai_mode: 0                # cross surface within Google

# --- Optimization Capacity (REQUIRED) ---
content_production_capacity_pages_per_month: 0
content_refresh_capacity_pages_per_month: 0
schema_rework_capacity_pages_per_month: 0
brand_voice_flexibility: ""                  # "single_locked_voice" | "two_modes_supported" | "per_audience_flexible"

# --- Substrate Prerequisite (REQUIRED, see framework-contentfirst.md) ---
contentfirst_score: 0                        # out of 30
substrate_in_first_byte: false
schema_in_first_byte: false
js_required_for_primary_content: false

# --- Foundational Authority (REQUIRED) ---
eeat_self_assessment_score: 0                # out of 130, see framework-eeat.md
infogain_score: 0                            # see framework-infogain.md
schema_core_graph_present: false             # see framework-schema.md
entitysalience_score: 0                      # see framework-entitysalience.md
knowledgegraph_status: ""                    # "wikidata_qid_assigned" | "wikipedia_article_live" | "neither" | "knowledge_panel_claimed"

# --- Per Engine Bot Access (REQUIRED) ---
gptbot_allowed: false
oai_searchbot_allowed: false
chatgpt_user_allowed: false
perplexitybot_allowed: false
perplexity_user_allowed: false
claudebot_allowed: false
google_extended_allowed: false
bingbot_allowed: false
meta_externalagent_allowed: false

# --- Tracking (REQUIRED) ---
multi_engine_citation_tracker: ""            # "Profound" | "Otterly" | "AthenaHQ" | "BrightEdge_AI_Catalyst" | "Semrush_AI_Toolkit" | "manual_only" | "none"
manual_sampling_cadence: ""                  # "weekly_priority_queries" | "monthly" | "quarterly"
server_log_per_engine_bot_monitoring: false
last_multi_engine_audit_date: ""

# --- Tradeoff Acknowledgment (REQUIRED, after Section 8) ---
tradeoff_documentation_complete: false       # signed off engagement document acknowledging engine priority order
tradeoff_review_cadence: ""                  # default "quarterly"

Citation tradeoff work cannot start until contentfirst_score is at least 22, eeat_self_assessment_score is at least 90, the core graph schema is server rendered, and the four engine specific frameworks have been read at minimum at the Section 5 (per engine ranking signals) level. Sites failing those dependencies route back to those frameworks first.

3. What the Tradeoff Question Actually Is

3.1 The "Win All Engines" Mirage

The first thing a multi engine optimization conversation needs to handle is the assumption that an undifferentiated AI strategy lifts every engine simultaneously. The data refutes that assumption decisively. The Yext analysis of 17.2 million citations across ChatGPT, Gemini, Claude, and Perplexity in Q4 2025 found 89 percent of cited domains were platform specific; only 11 percent of cited domains were shared across multiple engines. The Profound August 2024 to June 2025 analysis of cross platform citation patterns found a similar 11 percent overlap between ChatGPT and Perplexity. The Ahrefs February 2026 study of 730,000 Google response pairs found only 13.7 percent citation overlap between Google AI Overviews and AI Mode despite both surfaces running on the same Gemini 3 Pro engine since January 27, 2026.

A "win all engines" strategy means producing content and signals that satisfy the median of every engine's selection logic. The median is rarely the citation winner on any given engine. Engines that prefer institutional sources do not cite the median; engines that prefer Reddit and community discourse do not cite the median; engines that prefer YouTube transcripts do not cite the median. The result is broad mediocrity: cited intermittently on every engine, dominant on none.

The actionable insight from the divergence data: the engines are not interchangeable retrieval surfaces with different presentation. They are structurally different products with different bot families, different indices, different ranking signals, different source preferences, and different audiences. Treating them as one surface is the central anti pattern this framework exists to prevent.

3.2 What "Tradeoff" Means Operationally

A tradeoff is a decision where capacity allocated to one engine reduces capacity for another. The capacity dimensions in scope:

Capacity Dimension Engine A Choice Engine B Cost
Content production budget Long form research articles for Perplexity Fewer topical hub pages for ChatGPT
Schema rework hours Citation property and DefinedTerm for AI Citations breadth Less Article schema dateModified maintenance for AI Overviews
Freshness cadence Monthly refresh of 50 priority pages for Perplexity Less new evergreen content for topical authority hubs
Brand voice formality Institutional tone for Bing Copilot and Claude Less conversational tone for ChatGPT
Citation density Heavy primary source attribution per Perplexity Less narrative flow for ChatGPT topical hubs
Bot access posture Allow all AI bots maximally Lost control of training corpus exposure
Internal linking topology Hub and spoke for AI Overviews and topical authority Less individual page weight transfer for one off citations
Reddit and forum posture Active community participation for ChatGPT and Perplexity Team time off the website work itself
Wikipedia and Knowledge Graph work Months of Wikidata and Wikipedia citation work for Gemini and Claude Delayed individual content production

Every line above is a real engagement decision. The tradeoff document, produced from Section 7 and 8 work, captures which lines are accepted and which are deferred for the engagement.

3.3 The Choice Forced by Engine Specific Signal Weights

Engine specific signal weights forcing the tradeoff (synthesized across the Section 6 data, expanded per engine in Section 5):

Each engine's signal weights are coherent internally; across engines they conflict. The tradeoff document maps that conflict.

3.4 The Decision Stage Sequence

The decision sequence this framework runs:

  1. Establish ICP geographic split and business type (Section 2)
  2. Establish current per engine citation state (Section 2 and Section 13.1)
  3. Run the priority decision tree (Section 7)
  4. Document the per engine tradeoffs (Section 8)
  5. Deploy the universally additive unified strategy (Section 9)
  6. Deploy the priority engine specific divergent strategy (Section 10)
  7. Measure per engine lift (Section 13.5)
  8. Revisit the priority order quarterly (Section 14)

The order matters. Skipping to step 6 is the most common engagement failure mode: divergent strategy without unified foundation, which produces no citation lift on any engine because the substrate prerequisite is unmet.

4. The Engine Map

4.1 The Eight Engines In Scope for 2026

The cross engine surface decomposes into eight distinct engines as of Q2 2026. Each operates with a different bot family, index, ranking signal weight, and audience profile. The Section 5 matrix scores them on 12 signals; this section establishes who each engine is and which audience it reaches.

Google AI Overviews (AIO). The generative summary appearing above organic results on approximately 48 percent of all Google searches in Q1 2026 per multiple 2026 statistics aggregations. Runs on Gemini 3 Pro since January 27, 2026. Cites approximately 3 to 8 sources per response, mentions brands by name approximately 61 percent of the time, citation rate 84.9 percent. Audience: every Google searcher globally where the surface triggers. See framework-aioverviews.md.

Google AI Mode. The standalone conversational tab in Google search, also Gemini 3 Pro since January 27, 2026. 75 million daily active users in December 2025 per Google's Q3 2025 earnings disclosure and the etavrian Q3 analysis; queries doubled across the quarter to reach approximately 1 billion per month across 100 million US plus India active users. Cites 5 to 12 sources per response; citation rate 76.3 percent; brand mention rate 37.6 percent. 13.7 percent citation overlap with AI Overviews per the Ahrefs February 2026 study of 730,000 response pairs. Audience: Google users who explicitly switch to AI Mode for conversational research. See framework-aioverviews.md.

ChatGPT Search. The browsing capability inside ChatGPT (launched October 31, 2024 per OpenAI's announcement; ChatGPT itself reached approximately 800 million weekly active users in October 2025 per Sam Altman's TechCrunch interview and 900 million weekly by February 2026 per the TechCrunch update). Approximately 18 percent of ChatGPT conversations trigger at least one web search per the Profound 7,000 query analysis 2025. Estimated 250 to 500 million weekly queries through the retrieval layer per the Similarweb 2026 AI Search report. Citation rate per the BrightEdge 2026 weekly insights series approximately 56 percent of queries; 3 to 7 cited sources per response. Audience: ChatGPT's massive global user base, weighted toward the US per the 5W Q1 2026 index. See framework-searchgpt.md.

Perplexity. The answer engine with prominent inline citation display (launched December 7, 2022 per Wikipedia). Approximately 45 million MAU by early 2026 per the Panto AI 2026 statistics; approximately 1.2 to 1.5 billion queries per month per Wytlabs 2026. Average 8.79 citations per response per BrightEdge 2025 to 2026; citation rate 97 percent per the BrightEdge 2026 weekly insights series. Audience: research oriented users, often technical or academic, weighted toward US English language queries. See framework-perplexityspaces.md.

Claude with web search. Claude.ai's web search and web fetch integration, with citations inline per Anthropic's documentation. Citation rate approximately 39 percent of queries per the BrightEdge 2026 weekly insights series; 2.1 sources per cited response. Behavior varies sharply by query type: 95 percent citation rate for discovery queries; 0 percent for informational queries answered from training. Audience: technical and analytical users, knowledge workers, developers. See framework-aicitations.md.

Bing Copilot. Microsoft's Copilot in Bing chat, Edge, and Windows. Bing Webmaster Tools AI Performance dashboard public preview since February 9, 2026 surfaces Copilot citation data per the Bing webmaster blog announcement. Heavy weight on institutional sources and Bing index ranking. The Bing April 2026 algorithm updates formalized AI driven signals: authority, structured data accuracy, entity relevance. Audience: Windows users, enterprise Microsoft 365 deployments, Edge browser users. Smaller than ChatGPT's audience but commercially valuable. See framework-aicitations.md.

Meta AI. Meta's assistant inside Facebook, Instagram, WhatsApp, and Threads. Source preferences less documented than the major engines in independent studies; per the Pressonify 2026 platform comparison and the Authority Tech 2026 publications analysis, Meta AI weights social signals (Facebook, Instagram engagement) higher than competitors. Audience: Meta's social platform users. Marketing relevance is for consumer brands embedded in Meta's social commerce flow. See framework-aicitations.md.

Grok. xAI's Grok inside X (Twitter). Heavy preference for X content as a citation source per the broader AI citations framework analysis. Real time information weighted higher than other engines; less developed authority signals. Audience: X users and X integrated workflows. Marketing relevance is brand voice and crisis monitoring rather than primary citation acquisition. See framework-aicitations.md.

4.2 Audience Distribution and Query Mix

The audience and query mix per engine, synthesized from the 2025 to 2026 sources:

Engine Approximate Audience Query Mix Bias
Google AI Overviews Every Google searcher where AIO triggers (approximately 48 percent of searches in Q1 2026 per multiple sources) Broad. Informational, commercial, navigational; over 70 percent on informational and how to queries per the position.digital 2026 statistics roundup.
Google AI Mode 75 million DAU per Google Q3 2025 earnings; 100 million MAU US plus India Conversational research, exploratory, multi turn. Less navigational.
ChatGPT Search 900 million WAU ChatGPT total per TechCrunch February 2026; approximately 18 percent of conversations trigger Search per Profound 2025 Topical, conversational, broad commercial; biased toward US English per the 5W Q1 2026 index.
Perplexity 45 million MAU per Panto AI 2026 Research, technical, primary source seeking. Higher academic and professional weight than ChatGPT.
Claude with web search Anthropic does not publish DAU; Claude weekly active users have been described as in the tens of millions in 2026 industry reports Analytical, knowledge worker, technical. Discovery queries cite at 95 percent; informational at 0 percent per the BrightEdge 2026 analysis.
Bing Copilot Windows and Edge user base, enterprise Microsoft 365 Productivity, B2B, enterprise research; heavy institutional source weight.
Meta AI Meta's billions of social platform users Conversational inside social apps; consumer oriented.
Grok X user base Real time news, social commentary, niche communities on X.

The audience distribution determines which engine matters for a given business. A US English language B2B SaaS company with technical decision maker buyers cannot reach its ICP through Meta AI; a Brazilian consumer e commerce brand cannot reach its ICP through Claude with web search. Section 11 maps audience to engine priority.

4.3 The Concentration Pattern

Per the BrightEdge weekly AI search insights data 2025 to 2026, the top 10 most cited domains capture different shares of total citation surface per engine:

The concentration ratio is operationally important. Long tail engines (ChatGPT, AI Mode) reward sites with strong individual page authority on niche topics. Concentrated engines (Perplexity, AI Overviews) reward sites with institutional positioning or recognized authority. A long tail strategy that wins ChatGPT may not move Perplexity or AI Overview citation share at all because the candidate set is structurally tighter.

5. The Signal Weight Comparison Matrix

This is the central reference of the framework. The 12 signals below are scored as relative weight per engine on a 1 to 5 scale (1 = low weight, 5 = highest weight). The scores synthesize the Surfer SEO December 2025 study (173,902 URLs), Ahrefs February 2026 study (863,000 keywords; 730,000 response pairs), Profound 2025 to 2026 multi engine analyses, BrightEdge weekly insights 2025 to 2026, the 5W Public Relations Q1 2026 AI Platform Citation Source Index (680 million citations consolidated), Otterly January to February 2026 1 million plus citation analysis, the Yext Q4 2025 17.2 million citation analysis, the Whitehat SEO 2026 cross engine comparison, the Wellows 2025 Perplexity visibility study, and the Pressonify 2026 platform definitive guide.

5.1 The Matrix

Signal AI Overview AI Mode ChatGPT Search Perplexity Claude Web Bing Copilot Meta AI Grok
Server rendered substrate (first byte HTML) 5 5 5 5 5 5 4 4
Schema completeness and @graph pattern 5 5 4 4 4 5 3 2
E-E-A-T author and reviewer signals 5 4 4 4 5 5 3 2
Information Gain (novel contribution) 5 5 4 5 4 4 3 3
Primary source attribution on source page 3 3 3 5 4 4 3 2
Freshness within 30 days (dateModified) 4 4 5 5 3 4 3 5
Wikipedia and Wikidata entity presence 4 4 5 4 5 5 3 3
Reddit and community signal presence 2 3 5 5 3 2 2 3
YouTube transcript and video mention 5 4 3 3 2 2 3 2
Brand mention pace across the open web 4 4 5 4 4 4 4 4
Fan out query coverage and topical depth 5 5 3 4 3 4 3 2
Bing index presence (top 20 rank) 1 1 5 3 2 5 1 1

5.2 How to Read the Matrix

A signal weight of 5 means the engine routinely fails to cite a page that lacks the signal, regardless of other strengths. A signal weight of 4 means the signal is heavily preferred but not gating. A signal weight of 3 means the signal contributes but is not a primary driver. Signal weights of 2 and 1 mean the engine treats the signal as informational at best.

The matrix shows where engines genuinely converge and where they diverge:

5.3 Where the Matrix Lies

The matrix is a planning tool, not a published algorithm. Three caveats:

  1. Engine weights change. The Google AI Overviews to AI Mode migration to Gemini 3 Pro on January 27, 2026 shifted YouTube weight upward and top 10 organic correlation downward over a seven month window (76 percent to 38 percent per the Ahrefs February 2026 study).
  2. Query class matters. Perplexity weights primary sources heavily on research queries; for community queries it weights Reddit even more. The signal weight in the matrix is a query mix weighted average.
  3. The studies have geographic skew. Most published studies are US English language. EU, APAC, LATAM citation behavior may differ at the margins.

Treat the matrix as the orientation map. Use the Section 13.5 manual sampling and per engine trackers to validate per engagement.

5.4 Decision Rule From the Matrix

For each priority engine identified in Section 7, the Section 5.1 column reveals which signals are gating (5), preferred (4), and informational (3 and below). The optimization plan invests heavily in 5s and 4s for the priority engines, treats 3s as opportunistic, and ignores 1s and 2s unless they appear at 4 or 5 on a secondary priority engine.

For engagements optimizing across multiple priority engines, the union of 5 weighted signals across those engines becomes the required signal set. Where 5 weighted signals across engines do not conflict (substrate, E-E-A-T, schema, Information Gain), the work is universally additive (Section 9). Where they conflict (brand voice formality versus conversational tone, citation density versus narrative flow, Reddit posture versus institutional positioning), the engagement requires a divergent strategy decision (Section 10).

6. Citation Overlap Data

6.1 The Foundational Datasets

The cross engine overlap data this framework relies on:

The pattern is consistent across studies, sample sizes, methodologies, and dates: cited domain overlap across engines clusters at 11 to 14 percent. The 11 percent is the planning number this framework uses.

6.2 Pairwise Engine Overlap Matrix

Pairwise cited domain overlap, synthesized from the Section 6.1 sources:

Engine A Engine B Approximate Cited Domain Overlap Primary Source
Google AI Overview Google AI Mode 13.7 percent Ahrefs February 2026, 730,000 response pairs
ChatGPT Search Perplexity 11 percent Profound August 2024 to June 2025; Whitehat SEO 2026
Google AI Overview ChatGPT Search Approximately 12 percent Ahrefs February 2026 (12 percent of AI cited URLs rank in Google top 10); Passionfruit 15,000 query study
Google AI Overview Perplexity Approximately 11 to 14 percent Yext Q4 2025; Passionfruit 15,000 query study
ChatGPT Search Claude Web Not directly published; estimated 8 to 12 percent Inferred from Yext Q4 2025 four engine analysis
Perplexity Claude Web Not directly published; estimated 10 to 15 percent Inferred from Yext Q4 2025
ChatGPT Search Bing Copilot Approximately 30 percent Inferred from shared Bing index dependency per framework-searchgpt.md Section 6.1
Google AI Mode Perplexity Approximately 10 to 13 percent Yext Q4 2025; cross referenced with BrightEdge weekly insights 2026

The ChatGPT to Bing Copilot overlap is anomalously high because both engines retrieve from Bing's index as their primary real time retrieval layer per the Yoast 2026 and ClickRank 2026 ChatGPT Search analyses. This is the closest thing to a "two engine, one optimization" pairing in the matrix.

6.3 What 11 Percent Overlap Implies Strategically

The strategic consequences of 11 percent cross engine overlap:

  1. A page cited on engine A has approximately 1 in 9 probability of also being cited on engine B for the same query. The probability of multi engine citation is multiplicative across additional engines, so a page cited on three engines simultaneously is rare.
  2. The site domain matters less than per page signal alignment. A site can earn citation share on engine A and lose it on engine B for separate pages addressing different sub queries.
  3. Cross engine wins come from the universally additive signals (Section 9), not from engine specific signals. The 11 percent overlap is approximately the share of citations driven by universal signals; the 89 percent platform specific share is driven by engine specific signals (Section 10).
  4. Reporting requires per engine tracking. Aggregate "AI citations" reporting masks the 89 percent divergence and produces strategic blindness.
  5. The decision to optimize for multiple engines is a multiplicative cost, not an additive cost. Optimizing for two engines costs more than 2x; optimizing for four engines costs more than 4x, because each additional engine requires divergent strategy investment on top of the unified foundation.

6.4 Where the Overlap is Highest and Why It Matters

Two engine pairs have unusually high overlap and merit unified strategy treatment as a single target:

ChatGPT Search and Bing Copilot: approximately 30 percent overlap due to shared Bing index dependency. A site that ranks in Bing's top 20 for a query has a meaningful chance of being cited in both engines. The unified strategy: Bing Webmaster Tools verification, Bing IndexNow configuration, Bing search authority work. The new Bing Webmaster Tools AI Performance dashboard (public preview February 9, 2026) tracks Copilot citations and is the closest thing to a ChatGPT Search visibility proxy short of dedicated AI trackers.

Google AI Overviews and Google AI Mode: 13.7 percent overlap despite shared Gemini 3 Pro engine. Lower than the ChatGPT to Copilot pair because the two surfaces serve different query fan out patterns (8 to 12 sub queries for AIO; 9 to 16 sub queries for AI Mode) and apply different brand mention thresholds (61 percent for AIO; 37.6 percent for AI Mode). The unified strategy: comprehensive sub topic coverage to maximize the candidate retrieval surface across both fan out patterns. See framework-aioverviews.md Section 5.

All other engine pairs are sufficiently divergent that the unified strategy alone produces low cross engine citation lift; each requires its own divergent strategy investment.

7. The Optimization Priority Decision Tree

7.1 Inputs to the Decision

The priority order produced by this section depends on three inputs:

  1. ICP geographic split (Section 2 client variables): US dominant, US plus global, global non US, regional, single metro.
  2. Query type mix (per framework-keywordresearch.md): informational, commercial, navigational, transactional, local intent.
  3. Business type (Section 2 client variables; Section 11 detail): local service, B2B SaaS, e commerce, content publisher, YMYL vertical, professional service, media brand.

The decision tree below routes through these three inputs to produce a priority engine order of three engines (priority 1, 2, 3) plus secondary engines for monitoring.

7.2 Decision Tree

START
|
+-- Is the business US dominant ICP (>= 70 percent US revenue)?
|   |
|   +-- YES: continue to query type
|   |
|   +-- NO: route to global ICP branch (Section 7.3)
|
+-- Query type mix dominant?
|   |
|   +-- Informational or how to (>= 50 percent of priority queries)
|   |   --> Priority: Google AI Overview, ChatGPT Search, Perplexity
|   |   --> Secondary: Google AI Mode, Bing Copilot, Claude Web
|   |
|   +-- Commercial intent (>= 50 percent of priority queries)
|   |   --> Priority: Google AI Overview, ChatGPT Search, Google AI Mode
|   |   --> Secondary: Bing Copilot, Perplexity
|   |
|   +-- Research or analytical (>= 50 percent of priority queries)
|   |   --> Priority: Perplexity, Claude Web, Google AI Mode
|   |   --> Secondary: ChatGPT Search, Bing Copilot
|   |
|   +-- Local intent (>= 50 percent of priority queries, geo modifier present)
|   |   --> Priority: Google AI Overview, Google AI Mode, Bing Copilot
|   |   --> Secondary: ChatGPT Search
|   |   --> Apply framework-localseo.md alongside
|   |
|   +-- Navigational or branded (>= 50 percent of priority queries)
|   |   --> Priority: every engine where the brand has presence
|   |   --> Brand mention pace becomes primary work
|   |
|   +-- Mixed (no single class >= 50 percent)
|   |   --> Priority: Google AI Overview, ChatGPT Search, Perplexity
|   |   --> Secondary: Google AI Mode, Bing Copilot
|   |   --> Default priority for unfocused query mix
|
+-- Business type overlay (Section 11)
|   --> If YMYL, increase Claude Web priority (E-E-A-T weighted highest)
|   --> If B2B SaaS, increase Bing Copilot priority (institutional source preference)
|   --> If e commerce, increase Perplexity priority (research intent)
|   --> If local service, lock in Google AI Overview as priority 1
|
+-- Output: priority_engine_1, priority_engine_2, priority_engine_3, secondary_engines list

7.3 Global ICP Branch

For businesses with less than 70 percent US revenue, the priority order shifts:

7.4 The Three Engine Cap

The framework caps priority engines at three. The reason: per the Section 6 divergence data, optimizing for four or more engines as priorities requires divergent strategy investment that exceeds typical engagement capacity. A site committed to four engine priorities typically delivers mediocre divergent strategy work on all four. A site committed to three engine priorities can deliver focused divergent work and capture the universal lift on all eight.

Engagements with exceptional capacity (in house SEO team of 5+, content production capacity of 20+ pages per month, schema engineering team, dedicated digital PR) can expand to four or five engine priorities. The default cap is three.

7.5 The Secondary Engine Posture

Secondary engines are monitored but not actively optimized for divergently. The unified strategy (Section 9) produces some citation share on secondary engines passively; the engagement does not invest divergent strategy capacity in them. Quarterly Section 13.5 sampling on secondary engines surfaces unexpected lift or unexpected decline, which can trigger a Section 14 priority order revisit.

7.6 The Priority Order Document

The output of Section 7 is captured in a one page priority order document:

# Engine Priority Order

**Engagement**: {{BUSINESS_NAME}}
**Date**: {{ISO_DATE}}
**Next review**: {{ISO_DATE_PLUS_90_DAYS}}

## ICP
- Geographic split: {{US_SHARE}} US, {{EU_SHARE}} EU, {{APAC_SHARE}} APAC, {{OTHER_SHARE}} other
- Query type mix: {{INFORMATIONAL}} informational, {{COMMERCIAL}} commercial, {{RESEARCH}} research, {{LOCAL}} local
- Business type: {{BUSINESS_TYPE}}
- YMYL classification: {{YMYL_CLASSIFICATION}}

## Priority Engines (active divergent strategy investment)
1. {{ENGINE_1}} (per framework {{FRAMEWORK_REF}})
2. {{ENGINE_2}}
3. {{ENGINE_3}}

## Secondary Engines (monitoring only)
- {{ENGINE_4}}
- {{ENGINE_5}}
- {{ENGINE_6}}

## Tradeoffs Accepted (Section 8 references)
- Investment in {{ENGINE_1_REQUIREMENT}} reduces capacity for {{COST}}
- Investment in {{ENGINE_2_REQUIREMENT}} reduces capacity for {{COST}}

## Universal Strategy (Section 9, deployed across all engines)
- Substrate doctrine compliance per framework-contentfirst.md
- E-E-A-T pillars per framework-eeat.md
- Information Gain per framework-infogain.md
- Schema completeness per framework-schema.md
- Knowledge Graph per framework-knowledgegraph.md
- Brand mention pace per framework-digitalpr.md

## Divergent Strategy by Priority Engine (Section 10 details)
- {{ENGINE_1}}: {{SPECIFIC_DIVERGENT_WORK}}
- {{ENGINE_2}}: {{SPECIFIC_DIVERGENT_WORK}}
- {{ENGINE_3}}: {{SPECIFIC_DIVERGENT_WORK}}

## Sign Off
{{CLIENT_NAME}} {{DATE}}
{{AGENCY_LEAD}} {{DATE}}

The signed document is the artifact authorizing the engagement to deprioritize secondary engines and commit divergent investment to priority engines.

8. Engine Specific Tradeoffs

Per priority engine, what gets optimized for and what it costs.

8.1 Google AI Overviews and AI Mode Tradeoffs

What you optimize for:

What it costs elsewhere:

The hardest tradeoff: AIO and AI Mode share Gemini 3 Pro but cite different sources (13.7 percent overlap). Optimizing for AIO does not automatically lift AI Mode. The engagement either invests in both surfaces separately (doubling work) or picks one and accepts the citation gap on the other. The framework default for AIO priority engagements: optimize AIO primarily, monitor AI Mode quarterly, escalate AI Mode to divergent work only if AI Mode tracking shows commercial intent volume that justifies it.

8.2 ChatGPT Search Tradeoffs

What you optimize for:

What it costs elsewhere:

The hardest tradeoff: ChatGPT and Bing Copilot share approximately 30 percent overlap due to the Bing index dependency. Investment in ChatGPT Search work delivers free Bing Copilot lift, but only if the Bing Copilot priority weight justifies the dual investment. If the engagement prioritizes ChatGPT and treats Bing Copilot as secondary, the work is efficient. If the engagement prioritizes Perplexity and ChatGPT, the Bing infrastructure work does not transfer to Perplexity, and the engagement bears full ChatGPT cost.

8.3 Perplexity Tradeoffs

What you optimize for:

What it costs elsewhere:

The hardest tradeoff: Perplexity's primary source attribution signal is unique to Perplexity in the matrix (weight 5 versus 3 to 4 elsewhere). Pages optimized for Perplexity feel different from pages optimized for ChatGPT or AIO; the engagement either commits to the Perplexity voice on priority pages or sacrifices Perplexity citation rate. Pages built for ChatGPT first third density rarely earn Perplexity citation; pages built for Perplexity citation density underperform on ChatGPT.

8.4 Bing Copilot Tradeoffs

What you optimize for:

What it costs elsewhere:

The hardest tradeoff: Bing Copilot's audience is narrower than ChatGPT's or AI Overviews's. Investment in Copilot first work makes sense for B2B SaaS, enterprise software, and Microsoft 365 adjacent businesses; for consumer brands the audience overlap with consumer surfaces (ChatGPT, AIO) is the higher leverage path. Section 11 routes by business type.

8.5 Claude with Web Search Tradeoffs

What you optimize for:

What it costs elsewhere:

The hardest tradeoff: Claude is the most discerning engine on E-E-A-T but the lowest volume engine. The engagement decision: optimize for Claude only when E-E-A-T work is required anyway (YMYL verticals, B2B SaaS with technical buyers, knowledge worker audiences), so the marginal Claude investment is small on top of the necessary E-E-A-T work. For other engagements Claude becomes secondary monitoring rather than priority divergent investment.

9. Unified Strategy Patterns

The signals where investment lifts every engine. These are the Section 5 matrix rows that score 4 or 5 across the priority engine set for most engagements. Universal because they reflect what every engine's selection logic prefers, even when other signals diverge.

9.1 Substrate Quality (Universal Floor)

Server rendered substrate scores 5 across every engine in the matrix except Meta AI and Grok (which score 4 because their crawlers are less strictly non JS). No engine cites a page whose first byte HTML is empty. The substrate prerequisite is the universal floor; every engine specific framework gates citation work on contentfirst_score of at least 22 (framework-contentfirst.md).

Investment: server side rendering, static site generation, prerendering, or hybrid SSR/CSR with server rendered substrate. The triage test:

curl -A "Mozilla/5.0 AppleWebKit/537.36 (KHTML, like Gecko); compatible; GPTBot/1.2; +https://openai.com/gptbot" \
  -s https://example.com/priority-page/ | less

Repeat with PerplexityBot, ClaudeBot, and Bingbot user agents. The output must contain the H1, lede, all H2s, FAQ content as actual text, schema JSON-LD, author byline, dateModified, and internal links. Anything missing is invisible to the engine using that user agent.

Cross stack patterns for substrate compliance are in framework-cross-stack-implementation.md, framework-react.md, and framework-tailwind.md.

9.2 Schema Completeness (Universal High Weight)

Schema completeness scores 4 to 5 across every engine except Meta AI (3) and Grok (2). Schema is universal because it provides structured metadata every engine uses for entity disambiguation, regardless of how that engine weights other signals.

The required schema stack per page type:

The @graph pattern with @id cross references is preferred across every engine per the Section 5 matrix. Schema must be server rendered in the document head, not GTM injected. See framework-schema.md.

9.3 E-E-A-T Pillars (Universal for YMYL, Heavy for Non YMYL)

E-E-A-T author and reviewer signals score 4 to 5 across every engine except Meta AI (3) and Grok (2). Every engine that cares about citation quality cares about E-E-A-T. For YMYL sites the signal is gating; for non YMYL sites it is preferential but heavily weighted.

The required pillars:

See framework-eeat.md, framework-ymyl.md, framework-trustsignals.md.

9.4 Information Gain (Universal Differentiator)

Information Gain scores 4 to 5 across every engine. The signal is universal because every engine's synthesis layer is differential: it picks sources that contribute novel information rather than sources that paraphrase common knowledge.

The Section 5 Surfer SEO December 2025 finding: AIO cited articles cover 62 percent more facts than non cited articles. The Ahrefs February 2026 finding: AI Overviews cite AI generated content more than human writing in some categories, which is a Information Gain failure mode (the AI generated content scores higher on novel claim density per page even when factually thinner). The lesson: Information Gain is measured per page on factual density and novel contribution, not on whether the content is human written.

Investment per framework-infogain.md:

Information Gain is the single highest leverage universally additive signal for sites whose content currently aggregates rather than contributes.

9.5 Knowledge Graph Presence (Universal for Recognized Entities)

Wikipedia and Wikidata entity presence scores 4 to 5 across every engine except Meta AI and Grok (3 each). The Knowledge Graph signal is universal because every engine uses entity recognition to match query entities to candidate pages, and Wikipedia plus Wikidata is the canonical entity graph the engines have ingested at training time.

For notable entities (businesses, products, services, people) where the Wikipedia notability threshold and reliable source threshold can be met, Wikipedia plus Wikidata presence is the highest leverage universal investment. For non notable entities the work is Wikidata Q-ID only (which has a looser threshold) and per framework-knowledgegraph.md the entity graph work via Schema.org sameAs arrays.

9.6 Brand Mention Pace (Universal Compounding Signal)

Brand mention pace across the open web scores 4 across every engine. Brand mention pace compounds: every additional mention on a recognized industry publication, podcast, YouTube transcript, Reddit thread, or news article increases entity recognition probability across every engine simultaneously.

Investment per framework-digitalpr.md:

Brand mention pace is the highest leverage signal for sites that already pass the substrate, schema, E-E-A-T, and Information Gain floors. It is the marginal investment that distinguishes sites cited regularly from sites cited intermittently.

9.7 The Universal Strategy Deployment Order

Per Section 1.3, the universal strategy is deployed before any divergent strategy. The order:

  1. Substrate doctrine compliance (contentfirst_score 22+)
  2. Schema completeness with @graph pattern
  3. E-E-A-T pillars (with reviewer credit for YMYL)
  4. Information Gain per priority page
  5. Knowledge Graph presence (Wikidata Q-ID minimum; Wikipedia where notability supports)
  6. Brand mention pace pipeline

Engagements that complete the universal strategy first capture approximately 11 percent of citations across every engine (the cross engine overlap rate per Section 6.3). Engagements that skip directly to divergent strategy without completing the universal foundation typically capture no citation lift on any engine because the substrate prerequisite is unmet.

10. Divergent Strategy Patterns

The signals where engines genuinely require different optimization. These are the Section 5 matrix rows that diverge sharply across engines, where investment in one direction excludes investment in another.

10.1 Citation Density and Article Structure

Engines diverge on where in the article they extract content from.

Engine Preferred Citation Position Implication
ChatGPT Search First third of content (44 percent of citations) Top of article factual summary; 40 to 75 word answer paragraph after each H2
Perplexity Distributed citation density across the article Approximately one citable hyperlinked statistic per 150 to 200 words throughout
Google AI Overviews Sections of comprehensive sub topic coverage Fan out query coverage with sub topic depth per H2
Claude with web Analytical sections with clear methodology Methodology, sources, balanced perspective sections
Bing Copilot Institutional sources with named author attribution Visible author credentials at top; organizational credit

A page optimized for ChatGPT first third citation reads differently from a page optimized for Perplexity citation density. Sites with capacity to maintain two content patterns deploy both per page type (ChatGPT pattern for topical hub articles; Perplexity pattern for research pieces). Sites with single voice capacity choose based on Section 7 priority.

10.2 Brand Voice and Formality

Brand voice diverges across engines because each engine's training corpus and source preferences embed a different voice expectation.

Engine Preferred Voice Source
ChatGPT Search Conversational, accessible, factual Reddit and Wikipedia weighted sources skew toward natural language
Perplexity Research style, citation rich, primary source heavy Academic and government source preferences
Google AI Overviews Direct answer, factually dense, broad authority Brand controlled and authoritative content preference
Claude with web Analytical, balanced, methodologically clear Cautious citation behavior rewards analytical voice
Bing Copilot Institutional, organizationally credentialed Institutional source preference

A single locked brand voice forces a tradeoff. A B2B SaaS company committing to institutional voice for Bing Copilot citation accepts a ChatGPT citation gap on conversational queries. A consumer brand committing to conversational voice for ChatGPT accepts a Bing Copilot citation gap on enterprise queries. Engagements with brand voice flexibility (Section 2 client variable) maintain two modes; engagements without flexibility choose by Section 7 priority.

See framework-brandvoice.md for the voice operationalization layer.

10.3 Freshness Cadence

Engines weight freshness differently and reward different cadences.

Engine Freshness Sensitivity Cadence
ChatGPT Search 30 days primary; 2 year tail (Ahrefs 2026) Monthly priority pages; quarterly evergreen
Perplexity 30 day optimal (40 percent drop beyond, 65 percent beyond 90) Monthly priority pages; quarterly minimum
Google AI Overviews Variable by query type; informational query freshness matters less Quarterly priority pages; annual evergreen acceptable
Claude with web Less freshness sensitive than ChatGPT or Perplexity Quarterly priority pages
Bing Copilot Moderately freshness sensitive Monthly priority pages
Grok Heavy real time bias Daily or weekly for time sensitive content

A monthly refresh cadence for Perplexity priority engagements consumes significant content team capacity. A quarterly refresh cadence for AIO priority engagements is lower cost. The engagement either invests at the highest cadence required by the priority engine set or accepts citation share decline on the highest freshness engines.

See framework-contentrefresh.md for the refresh process layer.

10.4 Bot Access Posture

Engines control access through different bot families. The robots.txt posture diverges based on which engines are priority.

Bot Engine Priority Engagement Default
GPTBot ChatGPT training Deliberate decision per training opt out policy
OAI-SearchBot ChatGPT Search retrieval Allow if ChatGPT is priority
ChatGPT-User ChatGPT user invoked browsing Allow if ChatGPT is priority
PerplexityBot Perplexity index Allow if Perplexity is priority
Perplexity-User Perplexity live user fetch Allow if Perplexity is priority
ClaudeBot Claude training and web search Allow if Claude is priority
anthropic-ai Anthropic crawler Allow if Claude is priority
Google-Extended Google AI training Allow for AI Overview eligibility
GoogleOther Google other crawlers Allow universally
Googlebot Google search and AIO retrieval Allow universally
Bingbot Bing search and Copilot retrieval Allow if ChatGPT or Copilot priority
Meta-ExternalAgent Meta AI Allow if Meta AI priority
Diffbot Knowledge graph crawler Allow universally
CCBot Common Crawl Allow universally
Applebot-Extended Apple Intelligence Allow universally

The maximally permissive posture (allow all major AI bots) is the default for most engagements per framework-aicitations.md Section 6.1. Sites with deliberate training opt out policies make per bot decisions; the Section 7 priority engines determine which bots must be allowed regardless of training opt out.

10.5 Reddit and Community Signal Posture

Reddit posture diverges sharply by engine.

Engine Reddit Weight Recommended Posture
ChatGPT Search 5 (11.97 percent of US citations per 5W Q1 2026) Active organic subreddit participation
Perplexity 5 (6.6 to 24 percent depending on sample window) Active organic participation; brand mention monitoring
Google AI Mode 3 Monitor; no active investment unless community queries dominate
Google AI Overview 2 (Reddit 2.2 percent per Profound 2025) Monitor only
Claude with web 3 (Reviews and social media weight 2 to 4x competing models per Yext Q4 2025) Indirect via review platform presence
Bing Copilot 2 Monitor only; institutional sources weight higher
Meta AI 2 Monitor only
Grok 3 Monitor; X presence weight higher

ChatGPT and Perplexity priority engagements invest in organic Reddit posture. AIO and Bing Copilot priority engagements do not. The investment is team time on Reddit (organic engagement, long form comments with primary source links, no karma manipulation, no paid promotion, no AI generated comments per the warning in framework-perplexityspaces.md Section 5.9). The capacity cost is real; engagements without Reddit posture for community queries lose citation share to engagements with it.

10.6 YouTube and Video Mention Strategy

YouTube weight is divergent.

Engine YouTube Weight Investment
Google AI Overview 5 (18.2 percent of citations from outside top 100 per Ahrefs February 2026) High; pursue YouTube interviews, podcast appearances with video
Google AI Mode 4 High; same investment as AIO
ChatGPT Search 3 Moderate; transcripts ingested but lower weight than primary sources
Perplexity 3 Moderate; YouTube transcripts ingested
Claude with web 2 Low
Bing Copilot 2 Low
Meta AI 3 Moderate; social content adjacency
Grok 2 Low

The Ahrefs December 2025 correlation study cited in the position.digital 2026 AI SEO statistics roundup found YouTube mention correlation with AI search visibility at 0.737, the strongest single predictor measured. The investment is asymmetric: it pays off heavily on Google surfaces and AI Mode, modestly on ChatGPT and Perplexity, minimally on Bing Copilot and Claude. AIO and AI Mode priority engagements invest in YouTube creator partnerships and podcast appearances; other priority engagements treat YouTube as opportunistic.

10.7 The Divergent Strategy Investment Order

The deployment order per priority engine, after the universal strategy is in place:

  1. Apply the engine specific framework Section 5 ranking signals in full per priority engine.
  2. Apply the Section 5.1 matrix 5 weighted signals for each priority engine.
  3. Apply the Section 10 divergent patterns (citation density, brand voice, freshness, bot access, Reddit, YouTube) per priority engine.
  4. Validate per engine per Section 13.5 manual sampling and tracker data.
  5. Iterate quarterly per Section 14.

The framework cap of three priority engines (Section 7.4) is the practical limit because each additional priority engine requires its own divergent strategy investment on top of the universal foundation. Engagements that try to invest divergently in four or five engines typically deliver mediocre divergent work on all of them.

11. Engine Choice for Different Business Types

The Section 7 priority decision tree produces a starting point; the business type overlay refines it. Per business type, which engines actually drive the audience, which engines are commercially relevant, and which engines are wasted investment.

11.1 Local Service Businesses

Examples: contractors, mobile mechanics, pet waste removal, hardwood floor refinishing, cleaning services, lawn care.

Priority engines: Google AI Overview, Google AI Mode, Bing Copilot. Reasoning: local intent queries are dominated by Google surfaces; Bing Copilot picks up some local query share. ChatGPT Search and Perplexity are secondary because their query mix biases away from local intent.

Special considerations: GBP, Local Pack, and review platform work per framework-localseo.md is foundational. AIO and AI Mode citations for local services correlate with Local Pack presence and review quality.

Wasted investment: heavy Reddit posture, Wikipedia work (most local services do not meet notability threshold), YouTube creator partnerships (high cost low return for hyperlocal service businesses).

11.2 B2B SaaS

Examples: SaaS tools targeting technical decision makers, finance teams, marketing teams, ops teams.

Priority engines: Bing Copilot, Claude with web, Google AI Mode. Reasoning: B2B SaaS buyers research extensively in conversational AI tools; Microsoft 365 audience overlap drives Copilot relevance; technical buyers use Claude for analytical depth; AI Mode for exploratory commercial research. ChatGPT Search is secondary (high audience but lower commercial intent per the Profound 7,000 query analysis showing 18 percent web search trigger rate).

Special considerations: G2, Capterra, TrustRadius review platform presence per the Claude review signal weight. LinkedIn organic with named experts. Comparison content (vs competitors) for AI Mode commercial research. Per the averi.ai 2026 B2B SaaS citation benchmarks report, the engines diverge on what they recommend even when reaching similar conclusions; comparison content earns citation on multiple engines simultaneously.

Wasted investment: heavy Meta AI work (consumer audience mismatch), heavy Grok work (less developed B2B authority).

11.3 E commerce

Examples: physical product retailers, DTC brands, marketplace merchants, vertical e commerce.

Priority engines: Google AI Overview, ChatGPT Search, Perplexity. Reasoning: AIO drives commercial intent queries with Shopping graph integration; ChatGPT drives broad commercial research; Perplexity drives research oriented product comparison. Google AI Mode is secondary; Bing Copilot is secondary unless enterprise procurement audience is dominant.

Special considerations: Product, Offer, Review, AggregateRating schema per framework-ecommerceseo.md. MerchantReturnPolicy and shipping schema for AIO eligibility. Comparison content for cross brand commercial queries.

Wasted investment: heavy Claude work (lower citation rate per engine; technical audience mismatch for most consumer e commerce), heavy Wikipedia work (most products do not meet notability threshold).

11.4 Content Publishers

Examples: media brands, newsletter publishers, blogs at scale, niche content sites.

Priority engines: Google AI Overview, ChatGPT Search, Perplexity. Reasoning: content publishers have the substrate (well structured articles), the E-E-A-T (named authors and editorial process), and the Information Gain (original reporting where applicable) to compete on every major engine. The work is brand mention pace, Wikipedia and Wikidata presence for the publisher entity, and freshness cadence for time sensitive coverage.

Special considerations: News SEO per framework-newsseo.md for time sensitive content; Editorial policy and reviewer credit for YMYL adjacency.

Wasted investment: heavy Bing Copilot work unless the publisher targets enterprise audience specifically; heavy Meta AI work unless social adjacency is part of the publishing model.

11.5 YMYL Verticals (Healthcare, Finance, Legal)

Examples: clinics, medical practices, financial advisors, accountants, law firms, insurance brokers.

Priority engines: Google AI Overview, Claude with web, Bing Copilot. Reasoning: YMYL queries weight E-E-A-T heavily; Google AI Overviews has heavy YMYL guardrails; Claude is the most discerning engine on E-E-A-T per the Section 5 matrix (5 weight); Bing Copilot's institutional source preference aligns with credentialed YMYL content. ChatGPT Search and Perplexity are secondary but still relevant; the E-E-A-T threshold is high enough that pages built for YMYL Google AIO citation also tend to earn ChatGPT and Perplexity citation passively.

Special considerations: Reviewer credit mandatory per framework-ymyl.md. E-E-A-T self assessment score 110+. Vertical specific frameworks: framework-healthcare-seo.md, framework-finance-seo.md, framework-legal-seo.md.

Wasted investment: heavy Reddit posture for healthcare and legal where YMYL guardrails make community sourced answers risky; Meta AI work (consumer audience but YMYL guardrails limit citation surface).

11.6 Professional Services

Examples: agencies, consulting firms, freelancers with professional services, accounting firms not full YMYL.

Priority engines: Google AI Overview, ChatGPT Search, Bing Copilot. Reasoning: professional services have B2B audience overlap with B2B SaaS but smaller scale; AIO and Copilot drive business buyer queries; ChatGPT drives broad professional research.

Special considerations: LinkedIn organic presence; named partner and team bios with credentials; case study content for Information Gain; review and testimonial platform presence.

Wasted investment: heavy YouTube creator partnerships (high cost for typical professional services scale).

11.7 Media Brands

Examples: publications, news brands, podcast networks with publishing arms.

Priority engines: Google AI Overview, ChatGPT Search, Perplexity. Reasoning: media brands have the strongest natural substrate for citation across engines; the work is amplifying citation pace through brand mention work, Wikipedia presence for the brand entity, and freshness cadence.

Special considerations: News SEO; News sitemaps; editorial policy and reviewer credit; AggregateRating and Review schema where applicable.

Wasted investment: heavy Bing Copilot for consumer media brands; heavy Meta AI unless social distribution is the primary model.

11.8 The Business Type Decision Cross Check

After the Section 7 priority decision tree and the Section 11 business type overlay, the priority order is final. The cross check: do the priority engines collectively reach at least 80 percent of the ICP audience? If yes, the priority order is operationally sound. If no, expand to one additional priority engine to close the audience gap, accepting the capacity cost.

The audience math is approximate and per engagement; the Section 13.5 manual sampling and tracker data validate the audience fit quarterly per Section 14.

12. Common Multi Engine Mistakes (Top 10 Anti Patterns)

The ten anti patterns below cause cross engine citation failure even when individual engine optimization is competent. Addressing any three on a site that previously failed produces measurable lift within eight to twelve weeks.

12.1 Treating All Engines as One Surface

The central anti pattern. Treating ChatGPT, Perplexity, AI Overviews, and Claude as interchangeable retrieval surfaces and applying one optimization approach. Per Section 6, this produces approximately 11 percent citation share at best (the cross engine overlap rate) and silent invisibility on engine specific signal divergence.

Fix: run Section 7 priority decision tree. Document the priority order. Invest divergent strategy capacity per Section 10.

12.2 No Priority Order Document

A site optimizing for "AI" without explicitly naming priority engines and accepting their tradeoffs. Capacity gets allocated by whoever speaks loudest in engagement meetings rather than by strategic decision.

Fix: produce the Section 7.6 priority order document and get it signed off.

12.3 Universal Strategy Without Foundation Verification

Investing in Section 9 universal patterns without verifying the substrate prerequisite (contentfirst_score 22+), the schema graph, the E-E-A-T pillars, and the Knowledge Graph status. The universal strategy fails because the foundation is incomplete.

Fix: complete the Section 9.7 deployment order. Do not start divergent strategy until universal is verified.

12.4 Divergent Strategy Without Priority Order

The reverse failure mode. Investing in engine specific tactics (Reddit posture, YouTube partnerships, Bing IndexNow) without a priority order to validate the investment. Capacity gets spent on engines that do not drive the engagement's ICP audience.

Fix: complete Section 7 priority decision tree before any Section 10 divergent investment.

12.5 Optimizing for Four or More Engines as Priorities

Engagements that name four or five priority engines typically deliver mediocre divergent strategy work on all of them. Capacity is the binding constraint per Section 7.4.

Fix: cap at three priority engines. Demote the rest to secondary monitoring per Section 7.5.

12.6 Ignoring the Cross Surface Within Google

Treating Google AI Overviews and Google AI Mode as one surface because they share a Gemini engine. The 13.7 percent citation overlap per the Ahrefs February 2026 study means the surfaces are operationally distinct.

Fix: per framework-aioverviews.md, optimize for the surface that drives ICP audience volume. If AI Mode tracking shows commercial intent volume justifying the investment, escalate AI Mode to divergent work alongside AIO.

12.7 Aggregate AI Citation Reporting

Reporting "AI citations" as a single number across engines. The 89 percent platform specific share per Yext Q4 2025 means aggregate reporting masks per engine divergence. Engagements get blindsided by per engine declines because the aggregate number stays flat.

Fix: per engine citation tracking per Section 13.5. Aggregate is a secondary roll up, not the primary KPI.

12.8 Brand Voice Single Lock Without Awareness of Cost

Locking into a single brand voice (institutional or conversational) without acknowledging the per engine citation cost. Institutional voice loses ChatGPT citation; conversational voice loses Bing Copilot citation.

Fix: per Section 10.2, maintain two voice modes per page type if capacity allows. If capacity does not allow, document the tradeoff in the Section 7.6 priority order.

12.9 Freshness Cadence Single Standard

Applying one freshness cadence (typically quarterly) across all engines. Perplexity priority engagements lose 65 percent citation share beyond 90 days per the Wellows 2025 study; AIO priority engagements do not need monthly cadence.

Fix: per Section 10.3, cadence per priority engine. Monthly for Perplexity priority; quarterly for AIO priority; daily or weekly for time sensitive Grok content.

12.10 Reddit Posture for Engagements that Do Not Need It

Investing in organic Reddit engagement for AIO priority or Bing Copilot priority engagements where Reddit weight is 2 to 3 per Section 5. The capacity goes to community engagement time that does not lift the priority engines.

Fix: per Section 10.5, Reddit posture investment only for ChatGPT or Perplexity priority engagements. Other engagements monitor only.

13. Audit Rubric

13.1 Cross Engine Baseline Rubric (First 30 Days)

Run before any optimization investment to establish the baseline:

# Criterion Pass / Fail
B1 Priority order document complete and signed off (Section 7.6)
B2 Per engine citation status for priority queries documented (Section 2 variables)
B3 contentfirst_score 22+ (substrate prerequisite)
B4 eeat_self_assessment_score 90+
B5 Core graph schema server rendered (Organization, WebSite, WebPage, Person)
B6 Per priority engine bots allowed in robots.txt
B7 At least 30 days of server log data covering AI bot fetches
B8 Manual sampling baseline complete for priority query set across all priority engines
B9 Multi engine citation tracker selected and configured
B10 Tradeoff documentation acknowledging engine priority cost (Section 8)

Pass all 10 to proceed to Section 13.2. Failing any of B3 through B6 routes back to the substrate, E-E-A-T, schema, or bot access frameworks before continuing.

13.2 Per Page Audit Rubric (Priority Pages)

Per priority page across all priority engines:

# Criterion Pass / Fail
P1 curl test with multiple AI bot UAs returns H1, lede, H2s, FAQ, schema in first byte
P2 Lede 40 to 75 words, citable standalone
P3 Schema @graph with @id cross references, server rendered in head
P4 Article or BlogPosting schema with author, datePublished, dateModified
P5 Author byline with credentials and Person schema
P6 dateModified visible and in schema, refreshed within cadence per priority engine
P7 Information Gain demonstrated (original research, first hand, contrarian, expert interview, field data)
P8 Sub topic coverage addressing fan out queries (if AIO priority)
P9 Primary source attribution every 200 to 300 words (if Perplexity priority)
P10 First third of content carries factual summary (if ChatGPT priority)
P11 Analytical methodology section (if Claude priority)
P12 Institutional credentials and named experts (if Bing Copilot priority)
P13 Internal links use descriptive anchor text, three or more inbound
P14 Image alt text descriptive, not generic
P15 One citable hyperlinked statistic per 150 to 200 words

Score 15. World class: 13+ with zero P1 to P6 fails. P7 through P12 are gated by priority engine assignment from Section 7.

13.3 Site Wide Audit Rubric

# Criterion Pass / Fail
S1 Section 7 priority order document signed and current within 90 days
S2 Section 9 universal strategy fully deployed (substrate, schema, E-E-A-T, Information Gain, Knowledge Graph, brand mention pace)
S3 Section 10 divergent strategy deployed per priority engine
S4 Per priority engine bot access verified in robots.txt and server logs
S5 Per priority engine refresh cadence documented and operational
S6 Per priority engine tracker configured (Profound, Otterly, AthenaHQ, BrightEdge AI Catalyst, Semrush AI Toolkit)
S7 Manual sampling cadence per priority engine documented and running
S8 Cross engine citation overlap baselined (engine pairs from Section 6.2)
S9 Wikipedia and Wikidata posture aligned with priority engines requiring it
S10 Reddit posture aligned with Section 10.5 priority engine assignment
S11 YouTube and video strategy aligned with Section 10.6 priority engine assignment
S12 Brand voice mode count aligned with Section 10.2 priority engine count
S13 YMYL E-E-A-T compliance verified if YMYL classification is full or partial
S14 Per engine quarterly delta tracked and reviewed
S15 Tradeoff documentation reviewed within 90 days

Score 15. World class: 13+ with zero critical fails on S1 through S4.

13.4 First 90 Days Tracking

Day Range Action
0 to 14 Baseline: Section 13.1 rubric across all priority engines; priority order document signed
15 to 30 Section 9 universal strategy deployment for items not yet in place
31 to 60 Section 10 divergent strategy deployment per priority engine; verify per priority engine bot access in server logs
61 to 90 Re sample priority queries per engine; document deltas; identify next 50 query tier; first quarterly review of Section 7.6 priority order

13.5 Multi Engine Manual Sampling Protocol

Per priority query, sampled weekly:

  1. Open the priority engine's interface (chatgpt.com, perplexity.ai, claude.ai with web, gemini.google.com, copilot.microsoft.com, the Google AI Mode tab, the Google AI Overview area on a query that triggers it)
  2. Submit the priority query exactly as a user would
  3. Record whether the site is cited (yes / no)
  4. If cited, record position in the sources panel or citation list
  5. Record citation count for the response (benchmark per engine)
  6. Record brand mention rate (mentioned without citation)
  7. Repeat across the priority query set, across all priority engines

Aggregate into a per engine citation matrix per the Section 14.2 report template. Tooling: Profound for enterprise scale across all engines; Otterly for cost effective Perplexity and ChatGPT tracking; AthenaHQ for AI search analytics; BrightEdge AI Catalyst for enterprise SEO teams; Semrush AI Toolkit for AIO specifically. Manual sampling is feasible for sub 50 query lists.

13.6 Cross Engine Overlap Tracking

Metric Cadence Target
Per priority engine citation count from baseline Monthly Growing
Per engine pair overlap from baseline Quarterly Aligned with Section 6.2 expectations
Citation gain on priority engine 1 attributable to divergent strategy Quarterly Documented per page where possible
Universal strategy lift attributable to Section 9 deployment Quarterly At least 11 percent baseline cross engine share within first 90 days
Priority order revisit Quarterly Confirmed or updated with rationale

13.7 Implementation Report Template Outline

The implementation report covers: priority order rationale; universal strategy deployment status; divergent strategy deployment status per priority engine; baseline to current delta per priority engine; cross engine overlap baseline; tradeoff acknowledgment and review status. See Section 14.2 for the full template.

14. Maintenance Schedule and Report Templates

14.1 Maintenance Cadence

Weekly. Manual sample top 10 priority queries across all priority engines. Record per engine citation status, citation count, and brand mention. Check server logs for priority engine bots on priority pages. Refresh one priority page substantively if Perplexity is priority.

Monthly. Sample next 50 priority queries per priority engine. Schema validation sweep on priority pages. Audit robots.txt for priority engine bot allow lines. Review tracker dashboards per priority engine. Document deltas. Brand mention pace review per framework-digitalpr.md.

Quarterly. Site wide Section 13 audit. Revisit Section 7.6 priority order document; update if ICP, query mix, or business type shift. Refresh universal strategy compliance. Refresh divergent strategy per priority engine. Update Section 8 tradeoff documentation. Review secondary engine status for unexpected lift or decline.

Annually. Full framework review against current AI engine landscape. Refresh underlying engine specific frameworks (framework-aioverviews.md, framework-searchgpt.md, framework-perplexityspaces.md, framework-aicitations.md). Strategic review of priority engine selection and engine specific signal weights.

14.2 Implementation Report Template

# Multi Engine Tradeoffs Framework Implementation Report

**Site**: {{BUSINESS_NAME}}
**Date**: {{ISO_DATE}}
**Engagement lead**: {{LEAD_NAME}}

## Summary
- Priority engines: {{ENGINE_1}}, {{ENGINE_2}}, {{ENGINE_3}}
- Secondary engines: {{LIST}}
- Site wide audit score: X / 15
- Per page average: X / 15
- 90 day citation lift per priority engine

## Priority Order Rationale
- ICP geographic split: {{SPLIT}}
- Query type mix: {{MIX}}
- Business type: {{TYPE}}
- YMYL classification: {{CLASSIFICATION}}
- Section 7 decision tree output: {{ENGINE_PRIORITY_ORDER}}

## Universal Strategy Deployment
- Substrate doctrine: contentfirst_score before / after
- Schema completeness: before / after
- E-E-A-T pillars: score before / after
- Information Gain: pages before / after
- Knowledge Graph: Wikidata Q-ID, Wikipedia presence before / after
- Brand mention pace: mentions per quarter before / after

## Divergent Strategy by Priority Engine
For each priority engine:
- Section 10 patterns deployed
- Per page audit pass rate
- Citation count before / after
- Mention rate before / after

## Cross Engine Overlap
- Engine pair overlap baseline (Section 6.2)
- Current overlap per pair
- Universal strategy share vs divergent strategy share

## Tradeoffs Documented
- {{ENGINE_1}} investment cost to {{ENGINE_OTHER}}: documented and accepted
- {{ENGINE_2}} investment cost to {{ENGINE_OTHER}}: documented and accepted
- {{ENGINE_3}} investment cost to {{ENGINE_OTHER}}: documented and accepted

## Tracking
- Per engine tracker: {{TOOL}}
- Manual sampling cadence per engine
- Server log monitoring per engine bot

## Sign Off
{{CLIENT_NAME}} {{DATE}}
{{AGENCY_LEAD}} {{DATE}}

14.3 Audit Report Template

# Multi Engine Tradeoffs Framework Audit Report

**Site**: {{BUSINESS_NAME}}
**Date**: {{ISO_DATE}}
**Auditor**: {{AUDITOR_NAME}}

## Executive Summary
One paragraph assessment of multi engine posture.

**Site wide score**: X / 15
**Per page average**: X / 15
**Priority engine 1 citation rate**: X percent
**Priority engine 2 citation rate**: X percent
**Priority engine 3 citation rate**: X percent
**Cross engine overlap with peer engines**: X percent

## Findings by Section
- Section 7 priority order: present / absent, alignment with ICP and query mix
- Section 8 tradeoff documentation: present / absent
- Section 9 universal strategy: deployment status
- Section 10 divergent strategy: deployment per priority engine
- Section 11 business type alignment: aligned / misaligned
- Section 12 anti pattern presence: ten item check

## Critical Failures
List with remediation order.

## Recommended Remediation Order
Critical: priority order document, substrate, schema, E-E-A-T pillars
High: divergent strategy per priority engine, per engine bot access
Medium: cross engine overlap tracking, secondary engine monitoring, quarterly priority revisit cadence

## Sign Off
{{CLIENT_NAME}} {{DATE}}
{{AUDITOR}} {{DATE}}

14.4 Server Log Monitoring Bash Script

#!/usr/bin/env bash
# /var/www/sites/[domain]/scripts/multi-engine-bot-monthly-report.sh

LOG="/var/log/nginx/access.log"
ROLL="/var/log/nginx/access.log.1"
DOMAIN="$1"
DATE=$(date +%Y-%m-%d)
DIR="/var/www/sites/${DOMAIN}/reports/multi-engine-bots"
mkdir -p "${DIR}"
OUT="${DIR}/${DATE}.txt"

{
  echo "Multi Engine Bot Access Report - ${DOMAIN} - ${DATE}"
  echo ""
  echo "OpenAI family fetches:"
  echo "  GPTBot:"
  grep -c "GPTBot" "${LOG}" "${ROLL}" 2>/dev/null
  echo "  OAI-SearchBot:"
  grep -c "OAI-SearchBot" "${LOG}" "${ROLL}" 2>/dev/null
  echo "  ChatGPT-User:"
  grep -c "ChatGPT-User" "${LOG}" "${ROLL}" 2>/dev/null
  echo ""
  echo "Perplexity family fetches:"
  echo "  PerplexityBot:"
  grep -c "PerplexityBot" "${LOG}" "${ROLL}" 2>/dev/null
  echo "  Perplexity-User:"
  grep -c "Perplexity-User" "${LOG}" "${ROLL}" 2>/dev/null
  echo ""
  echo "Anthropic family fetches:"
  echo "  ClaudeBot:"
  grep -c "ClaudeBot" "${LOG}" "${ROLL}" 2>/dev/null
  echo "  anthropic-ai:"
  grep -c "anthropic-ai" "${LOG}" "${ROLL}" 2>/dev/null
  echo ""
  echo "Google family fetches:"
  echo "  Googlebot:"
  grep -c "Googlebot" "${LOG}" "${ROLL}" 2>/dev/null
  echo "  Google-Extended:"
  grep -c "Google-Extended" "${LOG}" "${ROLL}" 2>/dev/null
  echo ""
  echo "Microsoft family fetches:"
  echo "  Bingbot:"
  grep -c "Bingbot" "${LOG}" "${ROLL}" 2>/dev/null
  echo ""
  echo "Meta family fetches:"
  echo "  Meta-ExternalAgent:"
  grep -c "Meta-ExternalAgent" "${LOG}" "${ROLL}" 2>/dev/null
  echo ""
  echo "Common Crawl:"
  echo "  CCBot:"
  grep -c "CCBot" "${LOG}" "${ROLL}" 2>/dev/null
  echo ""
  echo "Apple Intelligence:"
  echo "  Applebot-Extended:"
  grep -c "Applebot-Extended" "${LOG}" "${ROLL}" 2>/dev/null
  echo ""
  echo "Non 200 status codes on AI bot requests (potential blocks):"
  grep -E "GPTBot|OAI-SearchBot|ChatGPT-User|PerplexityBot|Perplexity-User|ClaudeBot|anthropic-ai|Google-Extended|Bingbot|Meta-ExternalAgent|CCBot|Applebot-Extended" \
    "${LOG}" "${ROLL}" 2>/dev/null | \
    awk '$9 != 200 {print $9, $7}' | sort | uniq -c | sort -rn | head -25
  echo ""
  echo "Top paths by combined AI bot family:"
  grep -E "GPTBot|OAI-SearchBot|ChatGPT-User|PerplexityBot|Perplexity-User|ClaudeBot|anthropic-ai|Google-Extended|Bingbot|Meta-ExternalAgent" \
    "${LOG}" "${ROLL}" 2>/dev/null | \
    awk '{print $7}' | sort | uniq -c | sort -rn | head -25
} > "${OUT}"

echo "Report at ${OUT}"

Save at /var/www/sites/[domain]/scripts/multi-engine-bot-monthly-report.sh, chmod +x, cron monthly.

14.5 Cross Engine Quarterly Comparison Worksheet

For each priority query, sampled weekly and aggregated quarterly per priority engine:

Query AIO Q1 AIO Q2 Delta AI Mode Q1 AI Mode Q2 Delta ChatGPT Q1 ChatGPT Q2 Delta Perplexity Q1 Perplexity Q2 Delta Claude Q1 Claude Q2 Delta Copilot Q1 Copilot Q2 Delta Notes

Use to spot:

14.6 Tradeoff Review Cadence

The Section 7.6 priority order document is reviewed quarterly. Trigger criteria for unscheduled review:

Unscheduled reviews are documented in the Section 14.2 report as ad hoc reviews; quarterly reviews are documented as scheduled.

End of Framework Document

Document version: 1.0 Created: 2026-05-14 Maintained by: ThatDeveloperGuy

The multi engine citation surface is structurally divergent. Eleven percent inter engine cited domain overlap (Yext Q4 2025, 17.2 million citations across four engines; Profound August 2024 to June 2025, ChatGPT to Perplexity), 13.7 percent Google AI Overview to AI Mode overlap (Ahrefs February 2026, 730,000 response pairs), 89 percent platform specific share (Yext Q4 2025). A site cannot "win all engines" with a single optimization approach. The work is to pick three priority engines per ICP, query mix, and business type, deploy the universally additive signals (substrate, schema, E-E-A-T, Information Gain, Knowledge Graph, brand mention pace) to capture the 11 percent universal share, and invest divergent strategy capacity per priority engine to capture the 89 percent platform specific share where it matters.

The four engine specific frameworks (framework-aioverviews.md, framework-searchgpt.md, framework-perplexityspaces.md, framework-aicitations.md) specify what to do per engine. This framework specifies which engines and in what order. The substrate doctrine in framework-contentfirst.md gates every engine. Apply this framework after substrate and the universal strategy foundation are in place, alongside the four engine specific frameworks, with quarterly priority order review.

Companions

Want this framework implemented on your site?

ThatDevPro ships these frameworks as productized services. SDVOSB-certified veteran owned. Cassville, Missouri.

See Engine Optimization service ›