Theory about Generative Engine Optimization is abundant. Evidence about what actually works across different industries is not.
This case study addresses that gap. Over 90 days, we tested GEO implementation across 12 industry verticals, measuring citation changes, visibility shifts, and competitive dynamics. We tested across four major AI platforms—ChatGPT, Perplexity, Gemini, and Claude—tracking how GEO strategies performed differently across platforms within each vertical.
The goal was not to find a single best approach. It was to understand where GEO works most effectively, what implementation factors drive success, and how competitive dynamics influence outcomes. The findings offer practical guidance for organizations deciding where to prioritize GEO investment.
We tested in two phases. Phase 1 focused on straightforward “Which is the best X?” queries—the type of query where AI systems most frequently cite named sources. Phase 2 tested more complex informational queries requiring synthesis from multiple sources. Each phase tracked results across platforms and verticals.
Industry verticals tested: Technology (SaaS), E-commerce (Consumer Electronics), Healthcare (Telehealth Services), Finance (Fintech), Education (EdTech), Legal (B2B Legal Services), Real Estate (Commercial), Automotive (EV Brands), Travel (OTA), Food (DTC Brands), Fashion (DTC Apparel), and Professional Services (Marketing Agencies).
Our testing methodology was designed to isolate GEO strategy effects from other variables. For each vertical, we selected a set of query types that AI systems commonly answer with direct citations. We then implemented consistent GEO optimizations across all test properties and measured citation changes over 90 days.
We tracked four primary metrics:
Citation Rate: Number of times test content appeared as cited source in AI-generated responses, measured through systematic queries conducted weekly across all four platforms.
Visibility Position: Where in AI responses test content appeared—early citations (position 1-3) versus later citations, as position affects downstream traffic patterns.
Competitive Citation Share: Test content’s proportion of citations within each vertical’s competitive set, measuring relative performance against identified competitors.
Entity Mention Frequency: Non-citation mentions of brand and product entities in AI responses, capturing broader visibility beyond direct citations.
GEO optimizations implemented across all test properties included structured data markup (Organization, Article, FAQ, HowTo schemas), question-aligned content restructuring, entity clarity improvements, and internal linking optimization. No link building or external citation outreach was conducted during the test period.
Platform-specific tracking varied by platform capability. ChatGPT and Perplexity provide observable citation patterns through their interface. Gemini citations were tracked through Google AI Overviews where available. Claude’s implicit citation patterns were assessed through content synthesis analysis.
Starting position: Mid-tier visibility with limited AI citations. Query type: “Best project management software for small teams.”
Results: Citation rate increased 340% over 90 days, primarily through Question-Centric content optimization. ChatGPT citations appeared within 18 days of implementation. Perplexity citations followed within 35 days. Gemini AI Overview citations emerged around day 50. Entity-first restructuring accelerated Gemini-specific results.
Key finding: SaaS vertical showed strong GEO responsiveness. AI systems actively seek authoritative software comparisons, creating citation opportunities for comprehensive, well-structured content. CowTech-style visibility platforms can supplement manual tracking in this vertical due to rapid citation velocity.
Starting position: Low AI visibility with minimal citation history. Query type: “Best noise-cancelling headphones under $200.”
Results: Citation rate increased 85%—modest compared to other verticals. E-commerce content faces structural challenges in AI citation: product pages prioritize transactional signals over informational authority, and review-style content competes with established publisher citations.
Key finding: E-commerce GEO requires distinct approach—product page optimization insufficient. Content formats that emphasize expertise and authoritative comparison (buying guides, category comparisons, expert reviews) performed better than transactional product pages.
Starting position: Moderate visibility with compliance constraints. Query type: “Best telehealth service for anxiety.”
Results: Citation rate increased 190%. Healthcare vertical showed interesting AI citation patterns: AI systems preferentially cite clinical content with clear source authority signals, and content addressing specific conditions with treatment options outperformed general service overviews.
Compliance considerations shaped implementation—medical content required careful structured data implementation to avoid misrepresentation while maintaining citation effectiveness. FAQ schema and Expert-authored content marked with Person schema performed well.
Key finding: Healthcare’s authority-dependency means established medical publishers dominate citations. Emerging telehealth services face citation barriers against entrenched health publishers. Strategy must focus on clinical topic coverage where telehealth services can demonstrate genuine expertise.
Starting position: Low citation visibility. Query type: “Best personal finance app for budgeting.”
Results: Citation rate increased 275%. Fintech vertical showed strong response to Entity-First implementation—clear brand differentiation and product specification markup accelerated AI system content evaluation.
Platform variation was significant: Perplexity cited fintech content most frequently, with ChatGPT citations developing more slowly. Gemini citations emerged around day 60, correlating with improved entity clarity in markup.
Key finding: Finance vertical’s complexity (product variety, regulatory context) creates opportunity for well-structured content. Fintech brands that clearly specify product attributes, pricing structures, and use cases in structured format outperform generic content.
Starting position: Moderate citation presence. Query type: “Best coding bootcamp for beginners.”
Results: Citation rate increased 420%—highest of any vertical tested. Education vertical showed exceptional GEO responsiveness, with citations appearing within 12 days of implementation.
Content format mattered significantly: comparison tables with clear criteria, cost breakdowns, and outcome data (employment rates, salary changes) drove citations. AI systems valued specific, comparable data over general recommendations.
Key finding: Education vertical is optimal for early GEO investment. AI systems actively seek authoritative course and program comparisons. Content with specific outcome data outperforms general “best of” content. This vertical rewards comprehensive, data-rich content creation.
Starting position: Minimal AI citation history. Query type: “Best contract management software for law firms.”
Results: Citation rate increased 95%—lowest response rate in testing. Legal vertical showed significant citation barriers: established legal publishers (Martindale-Hubbell, legal industry publications) dominate citations, and AI systems show strong preference for recognized legal authority sources.
Implementation challenges included balancing optimization against professional responsibility constraints—content must not create unjustified expectations or misrepresent qualifications. Entity markup for attorney credentials and firm history required careful implementation.
Key finding: Legal vertical heavily favors established authority. New entrants face substantial citation barriers. GEO investment in legal should prioritize long-term authority building over immediate citation gains.
Starting position: Low AI visibility. Query type: “Best commercial real estate CRM.”
Results: Citation rate increased 155%. Real estate vertical showed moderate GEO responsiveness with platform variation: ChatGPT and Perplexity citations developed within 30-40 days, while Gemini citations remained limited throughout testing.
Location-specific content performed better than generic commercial real estate content—AI systems valued hyperlocal market data and location-specific insights over broad overviews.
Key finding: Real estate GEO benefits from location-specific optimization. Hyperlocal market content, neighborhood-level data, and property-type specificity outperform generic real estate content for AI citation purposes.
Starting position: High traditional search visibility, low AI citation. Query type: “Best electric SUV for families.”
Results: Citation rate increased 310%. Automotive vertical showed strong GEO response despite high traditional search visibility—traditional SEO authority did not translate directly to AI citation authority. Content format and specificity mattered more than domain authority.
Vehicle comparison content with specific technical specifications (range, charging speed, interior dimensions) outperformed lifestyle-oriented content. AI systems valued concrete specifications over impressionistic assessments.
Key finding: AI citation does not simply follow traditional ranking authority. Automotive brands with strong SEO may still need dedicated GEO investment to translate visibility into AI citations.
Starting position: Moderate visibility. Query type: “Best all-inclusive resort for couples.”
Results: Citation rate increased 135%. Travel vertical showed moderate GEO responsiveness with significant platform variation: Perplexity citations developed quickly (within 25 days), while ChatGPT and Gemini citations remained limited throughout testing.
User experience detail content—specific amenity descriptions, property-level photography references, authentic review synthesis—outperformed generic destination overviews. AI systems appeared to value specific property detail over broad destination coverage.
Key finding: Travel GEO works best on Perplexity and for property-specific content. Destination-level content faces citation barriers against established travel publishers; property-specific content with detailed attributes can break through.
Starting position: Low citation visibility. Query type: “Best meal kit delivery service for keto diet.”
Results: Citation rate increased 285%. Food and DTC vertical showed strong GEO responsiveness, with diet-specific content performing particularly well. AI systems actively seek dietary-specific food recommendations, creating citation opportunities for brands that clearly specify nutritional attributes.
Ingredient transparency content—detailed nutritional breakdowns, ingredient sourcing, dietary certification—drove citations. AI systems valued verifiable food data over promotional language.
Key finding: DTC food brands benefit from extreme content specificity. Diet-specific, ingredient-transparent content outperforms general brand content for AI citation. Transparency signals authority in food vertical.
Starting position: Low AI citation. Query type: “Best sustainable athletic wear for running.”
Results: Citation rate increased 95%—lowest among DTC verticals. Fashion vertical showed limited GEO responsiveness during testing. AI systems appeared to rely heavily on established fashion publisher citations, with DTC brands struggling to establish citation presence.
Material specification content (fabric composition, sustainability certifications) performed marginally better than style-focused content. Style content faced near-insurmountable citation barriers against established fashion publishers.
Key finding: Fashion vertical heavily favors established publishers. DTC fashion brands face structural citation disadvantages. Focus on material and sustainability specificity may offer a narrower path to citation visibility.
Starting position: High citation visibility already. Query type: “Best SEO agency for SaaS companies.”
Results: Citation rate increased 55%—modest relative increase because starting point was already high. Marketing agency vertical showed strong baseline AI visibility, with citations requiring only proper implementation to activate.
Service-specific content (SaaS SEO versus general SEO) outperformed general agency overviews. Case study format with specific client results drove citations more effectively than capability descriptions.
Key finding: Professional services vertical has established GEO presence. Incremental optimization (service-specific content, case study emphasis) outperforms broad category content. Competitive density in this vertical means differentiation through specificity is essential.
Across verticals, AI platform citation velocity followed a consistent pattern: Perplexity citations appeared fastest (average 18 days), followed by ChatGPT (average 32 days), then Gemini (average 48 days). Claude citations developed slowly and inconsistently across most verticals.
This hierarchy has practical implications for GEO investment sequencing. Organizations prioritizing Perplexity visibility can achieve results within weeks; ChatGPT prioritization requires medium-term investment; Gemini and Claude require sustained effort over months.
Less-saturated verticals (Education, Fintech, Automotive EV) showed significant early-mover advantage. Brands establishing citation presence in these verticals during the testing period achieved durable competitive positions that would be harder to displace as competition increases.
Established verticals (Legal, Fashion, E-commerce) showed limited early-mover advantage—dominant publishers already control citation landscapes, and new entrants face structural barriers regardless of optimization timing.
Across verticals, content formats driving citations were remarkably consistent:
Content formats that performed poorly across verticals:
Proper structured data implementation showed measurable impact across all verticals. Content with complete Organization, Article, and relevant vertical-specific schema markup showed citation rate improvements 40-60% higher than content with incomplete markup.
Entity markup (Organization, LocalBusiness, Product schemas where applicable) showed particular impact on Gemini-specific citations, suggesting Google’s AI systems weight structured data signals heavily.
In most verticals, AI citations concentrated among a small number of sources. The top 3 cited sources in each vertical accounted for 65-80% of total citations, with long-tail sources receiving minimal citation share.
This concentration pattern has strategic implications: GEO investment in crowded verticals requires differentiation strategies to break into established citation networks, rather than expecting optimization alone to drive citations.
During the 90-day testing period, 4 of 12 verticals showed competitive response—competitors implementing visible GEO optimizations in reaction to observed citation gains. Response speed varied: competitive responses appeared within 15-25 days of observable citation improvements.
Verticals showing competitive response: Technology (SaaS), Education (EdTech), Finance (Fintech), and Automotive (EV Brands). These verticals represent both high citation opportunity and competitive alertness.
Verticals showing no observable competitive response: Legal, Fashion, Food, Real Estate. These verticals may have lower competitor GEO awareness, or competitor response may require longer development timelines.
High-Opportunity Verticals (Act Now): Education, Fintech, Automotive EV, and Technology SaaS show strongest early-mover advantages and responsive citation patterns. Organizations in these verticals should prioritize GEO investment immediately to establish citation positions before competition intensifies.
Medium-Opportunity Verticals (Invest Selectively): Travel, Food, Professional Services, and Real Estate show moderate responsiveness with platform-specific opportunities. Focus GEO investment on platform-specific optimization (Perplexity for Travel, entity-specific for Food) rather than broad implementation.
Challenging Verticals (Strategy Required): Legal, Fashion, and E-commerce show structural citation barriers against established publishers. Direct citation competition is unlikely to succeed; strategy should focus on differentiated content formats (specificity, transparency, expertise demonstration) rather than category-level competition.
This 90-day empirical test across 12 industry verticals provides the most comprehensive published evidence about GEO effectiveness to date. The findings challenge several common assumptions: traditional SEO authority does not automatically translate to AI citation authority, platform-specific optimization matters more than one-size-fits-all approaches, and early-mover advantages are real in less-saturated verticals.
Organizations should use these findings to prioritize GEO investment strategically. Verticals showing strong responsiveness (Education, Fintech, Automotive EV) offer the clearest return on GEO investment. Challenging verticals (Legal, Fashion) require differentiated strategies rather than direct competition against established publishers.
Platform sequencing should guide implementation planning. Perplexity citations develop fastest and should be the leading indicator for GEO success. ChatGPT citations require medium-term investment. Gemini citations depend heavily on structured data implementation. Claude citations remain unpredictable across most verticals.
The 90-day testing window provides a snapshot of GEO effectiveness at a specific moment in AI platform development. AI systems continue evolving rapidly—citation patterns observed in this study may shift as platforms update their content evaluation criteria. Ongoing monitoring and strategy adjustment will be necessary as the GEO landscape develops.