> RESEARCH_STUDY_02

They Didn't Get Smarter.
They Made 4 Structural Changes.

160 AI responses. 110 cited brands. Every one audited. The pattern is structural — and replicable.

// How the Data Was Collected

In February 2026, 40 real industry queries were submitted across ChatGPT, Claude, Gemini, and Perplexity — four verticals, four platforms, 160 responses. Every cited brand was audited. No surveys. No proxies. Direct observation of what 110 cited brand websites had in common.

160
AI Responses Analyzed
4
AI Platforms Tested
110
Cited Brands Audited
4
Industry Verticals

// The Numbers That Define the Gap

88%
Responses Cited Brands
35.67x
Schema Markup Lift
99.1%
Had Reviews
2.5%
Cross-Platform Agreement

// The 6 Structural Differences

FINDING 01

Schema Markup: 35.67x Citation Lift, 97.3% Prevalence 35.67x LIFT

97.3% of cited brands had Schema.org structured data. Brands with Schema were cited 35.67 times more often than brands without it. Only 3 of 110 cited brands lacked Schema—all were small local businesses cited solely by Perplexity through live web search.

97.3% with Schema

What this means: Without Schema.org JSON-LD, you are structurally invisible to the citation layer of AI answer engines. Three exceptions exist across 110 brands — all were Perplexity live-search citations of local businesses. For everyone else, Schema is the entry requirement. See our Schema Markup for AEO guide →

FINDING 02

Reviews: Present in 99.1% of Cited Brands 99.1%

109 of 110 cited brands had customer reviews or testimonials visible on their site or on third-party platforms. The single exception was a small law firm cited once by Perplexity. AI engines use review presence as a trust and social proof signal when deciding which brands to recommend.

99.1% had reviews

Implication: Actively collect and display customer reviews. Google reviews, testimonials on your site, and third-party review platforms all count.

FINDING 03

FAQ Content: 63.6% Adoption, 1.75x Citation Frequency 1.75x LIFT

63.6% of cited brands had dedicated FAQ pages or FAQ sections with structured FAQPage schema. Brands with FAQ content were cited 1.75 times more often than those without. FAQ content gives AI engines pre-formatted Q&A pairs that are easy to extract and cite directly.

63.6% had FAQ pages
FINDING 04

Domain Authority: 91.8% — A Trailing Indicator, Not a Starting Point 91.8%

91.8% of cited brands had high domain authority scores. AI engines heavily weight established, credible domains. However, this is a trailing indicator—you build domain authority through the other signals (Schema, reviews, quality content), not independently.

91.8% high domain authority
FINDING 05

Cross-Platform Agreement: 2.5% — Each Engine Is a Separate Ranking System

All four AI platforms agreed on the top-cited brand in only 2.5% of queries (1 out of 40). Each engine has its own data sources, ranking logic, and preferences. This means optimizing for just one AI engine is insufficient—you need a strategy that works across all four.

Key insight: Claude cited brands 97.5% of the time (highest), but Perplexity was the only platform providing clickable source URLs. ChatGPT and Gemini mention brands by name without linking. Each engine delivers value differently.

FINDING 06

B2B Gets Named. Local Gets Directored.

B2B SaaS had the highest citation rate at 95%, with AI engines citing specific brand names and linking to service pages. Local service queries (HVAC, plumbing, legal) were different: AI engines primarily cited directory platforms like Yelp, Google Maps, and Angi (48.2% of all citations) rather than individual business websites.

For local businesses: Being listed and well-reviewed on major directories is as important as optimizing your own site. For B2B: your own site's AEO matters most.

// The Citation Priority Stack

This is the priority sequence. As you implement, start at Tier 1 and confirm each prerequisite before moving to accelerators and differentiators:

TIER 1: PREREQUISITES (Without these, you won't get cited)

Schema.org JSON-LD markup (97.3%) • Customer reviews (99.1%) • High domain authority (91.8%)

TIER 2: ACCELERATORS (These increase citation frequency)

FAQ pages with FAQPage schema (63.6%, 1.75x lift) • Pricing transparency • AI crawler access (robots.txt)

TIER 3: DIFFERENTIATORS (These determine who gets cited over competitors)

GIST algorithm alignment • Information gain density • Multi-engine optimization • llms.txt / ai.txt

// Cross-Study Validation

In our first case study, AEOfix achieved 70% AI visibility in 6 days using comprehensive Schema markup, FAQ-first content, and E-E-A-T signals. This 110-brand audit independently confirms that exact approach: the top three signals we implemented (Schema, FAQ content, authority signals) are the same three signals that 88% of all AI-cited brands share.

CASE STUDY 1: AEOfix

  • ✓ 9 Schema types implemented
  • ✓ FAQ-first content strategy
  • ✓ E-E-A-T authority signals
  • ✓ 70% visibility in 6 days

CASE STUDY 2: 110 BRANDS

  • ✓ 97.3% had Schema markup
  • ✓ 63.6% had FAQ content
  • ✓ 99.1% had review/trust signals
  • ✓ 88.1% citation rate across engines

The Complete Dataset Is Available

160 responses, per-engine breakdowns, industry analysis, and raw AI response data. As you review the full study, you'll have the complete picture behind every statistic on this page. The Source Map Report applies the same methodology to your brand — so you know exactly where you currently stand.

Read the Full Study Get Your Source Map — $59
WB
William Bouch
AEO Architect & Founder of AEOfix. Former construction worker turned full-stack developer. Engineering-driven AI visibility optimization.