What Pharma Brands Reveal About AI Search: Insights from komm.passion | Team Farner Citation Study - AI Search Visibility Blog | Insights and Data

Our German agency partner komm.passion | Team Farner just published something interesting. They analyzed how the five highest-grossing pharmaceutical products appear across AI search platforms in Germany, Austria, and Switzerland. They used OtterlyAI to track citations across ChatGPT, Gemini, Google AI Overview, and Perplexity.

The results confirm what we’ve been seeing in other industries, with a few pharma-specific twists worth your attention.

Check out the full (german) research results here.

AI Search Citation Study: How Pharma Brands Rank Across ChatGPT, Perplexity, and Google AI Overviews

A new study from German healthcare communications agency komm.passion | Team Farner analyzed website citations across four major AI search platforms to understand which websites get cited most frequently by AI models. The research examined the five highest-grossing pharmaceutical products (Keytruda, Ozempic, Dupixent, Biktarvy, and Eliquis) across Germany, Austria, and Switzerland using OtterlyAI’s citation tracking and overall AI Search monitoring tool.

The study analyzed responses from ChatGPT, Gemini, Google AI Overview, and Perplexity to 30 healthcare-related queries per product, totaling 150 questions across the DACH region. The findings reveal clear patterns about which content types, domain authorities, and optimization strategies drive AI citations in 2025.

Key Findings Summary

Citation Concentration

Top 10 websites capture approximately 50% of all AI citations
Remaining 50% distributes across 40+ additional domains
Regulatory agencies, medical journals, and established health portals dominate top citations

Geographic Patterns

German domains account for 40-60% of citations in Austria and Switzerland
Local domains (Austrian, Swiss) represent only 5-10% of citations in their markets
Cross-border citation is largely unidirectional (German → AT/CH, but not AT → CH)

Commercial vs. Official Content

High consumer demand products: 40-70% commercial/patient content citations
Specialized prescription products: 50-60% regulatory/official sources
Strong SEO investment translates directly to AI citation frequency
Commercial providers captured 3 of top 10 positions for Ozempic

Patient-Focused Content Performance

Patient websites achieved 40% of top-10 results for some queries
Patient content cited even for healthcare professional queries
Comprehensive, accessible content outperforms technical documentation
Industry-sponsored patient sites receive significant citation rates

Technical Requirements

Content behind paywalls or login walls receives minimal citations
PDF-heavy sites underperform compared to web-accessible text
Clear site structure and semantic markup improve citation likelihood
AI crawler access must be explicitly allowed in robots.txt

Study Methodology: How the Research Was Conducted

komm.passion | Team Farner developed 30 healthcare professional questions about five pharmaceutical products: Keytruda (oncology), Ozempic (diabetes/weight management), Dupixent (atopic dermatitis/asthma), Biktarvy (HIV therapy), and Eliquis (anticoagulation). These products represent the highest revenue pharmaceutical products globally in 2024-2025.

For each query, researchers extracted the top 50 cited sources from ChatGPT, Gemini, Google AI Overview, and Perplexity responses. All queries were conducted in German across three countries: Germany (DE), Austria (AT), and Switzerland (CH). The research team then clustered these sources by domain type, geographic origin, and content category to identify citation patterns.

OtterlyAI’s platform tracked which websites appeared most frequently, how often regulatory versus commercial content was cited, and whether local or regional domains dominated results. The total dataset included more than 1,500 individual citations analyzed across multiple dimensions.

Key Finding: Citation Concentration in AI Search Results

How Many Websites Dominate AI Citations?

The top 10 websites accounted for approximately 50% of all citations in the top-50 results across AI platforms. This concentration pattern appeared consistently across all pharmaceutical products analyzed, indicating that AI search engines strongly favor established, authoritative sources.

The most frequently cited sources included:

Regulatory agencies: G-BA (Gemeinsamer Bundesausschuss), IQWiG (Institute for Quality and Efficiency in Health Care)
Medical journals: Ärztezeitung, Pharmazeutische Zeitung
Health authorities: EMA (European Medicines Agency), national health agencies
Patient information portals: established health websites with comprehensive drug information

This 50% citation share means that breaking into the top tier of AI-cited sources requires substantial domain authority, content depth, and technical optimization. The remaining 50% of citations were distributed across 40+ additional domains, demonstrating steep competition for visibility outside the dominant group.

What This Means for Brand Visibility

AI search does not democratize information access; it concentrates citations among a small number of trusted sources. Brands not appearing in the top 10-15 most-cited domains for their category face significantly reduced visibility in AI-generated responses. This pattern mirrors but exceeds the concentration seen in traditional search engine results pages (SERPs).

Geographic Citation Patterns: How German Domains Dominate the DACH Region

What Percentage of AI Citations Come from German Websites?

German websites accounted for 40-60% of AI citations in Austria and Switzerland, despite queries being conducted in those countries. Local domains from Austria and Switzerland represented only 5-10% of citations in their respective markets. This means users in Zurich or Vienna see predominantly German sources when asking AI models about pharmaceutical products in German.

Specific examples:

Ozempic in Switzerland: 12 of 50 top results (24%) came from Swiss domains, while German domains dominated the remaining citations
Dupixent in Austria: 9 of 50 top results (18%) were Austrian domains
Eliquis in Switzerland: Multiple Swiss domains appeared in the top 10, including Eliquis.ch itself

Cross-Border Citation Patterns

Austrian sources rarely appeared in Swiss AI results, and Swiss sources rarely appeared in Austrian results. However, German sources appeared consistently across all three countries. This unidirectional influence means German content strategy has outsized impact across the DACH region.

For international brands, this finding suggests:

German-language content serves as the foundation for multi-country visibility
Well-optimized German websites will surface in neighboring markets without localization
Local market content still matters for top-10 placement in specific countries
Regional health portals and local authorities can achieve strong country-specific citations

When Local Content Wins

Local domains achieved strong performance in specific cases:

Country-specific product websites (example: Eliquis.ch in Swiss results)
National health portals with established domain authority
Local patient organizations with comprehensive information
Regional medical institutions and university hospitals

The pattern: local content secures top-tier placement within its home country but rarely crosses borders, while German content crosses borders consistently.

Commercial Content Performance in AI Search

Do Commercial Websites Get Cited by AI Models?

Yes. Commercial websites with strong SEO receive frequent citations from AI models, particularly for products with consumer demand and commercial competition. The Ozempic study demonstrated this pattern clearly.

Ozempic Citation Breakdown (Germany):

44 of 50 top results (88%) were German domains
3 of the top 10 sites were telehealth providers offering online Ozempic prescriptions
Professional medical journals: 0-6% of top-10 citations
Patient portals and commercial providers: 70%+ of top-10 citations
Online pharmacies, telemedicine platforms, and weight management services: heavily represented

Which Content Types Win Commercial Categories?

Content Type	Percentage of Top-50 Citations	Percentage of Top-10 Citations
Commercial/Pharmacy	31%	46%
Patient Portals	27%	14%
Regulatory/Associations	22%	6%
Professional Publications	6%	0%
Academic/Research	7%	14%
Other	13%	20%

The data shows commercial content performs exceptionally well when:

Strong consumer demand exists for the product
Commercial providers invest heavily in content marketing and SEO
The topic generates broad media discussion (example: Ozempic’s “weight loss injection” coverage)
Multiple providers compete for the same keywords

Why Commercial Content Dominates Certain Queries

AI models cite commercially optimized content because these sites:

Publish comprehensive, frequently updated information
Answer common user questions in accessible language
Maintain strong technical SEO and site architecture
Generate high engagement metrics and backlinks
Create content specifically designed for search visibility

Traditional medical journals and academic sources often use technical language, restrict access behind paywalls, or publish less frequently. Commercial providers fill the information gap with accessible, optimized content that AI models readily cite.

Patient-Focused Content Authority in AI Citations

How Often Do AI Models Cite Patient Information Websites?

Patient-focused websites appeared more frequently than expected in AI citations, even for prescription-only medications with healthcare professional target audiences.

Biktarvy Example (Austria):

4 of 10 top results (40%) focused on patient information
Industry-sponsored patient information sites were included
Patient organizations appeared alongside regulatory sources
Patient forums and community resources received citations

This pattern repeated across multiple pharmaceutical products. Patient organizations, patient information services, and patient advocacy websites consistently appeared in top-tier citations across Keytruda, Dupixent, and Eliquis queries.

Why AI Models Favor Patient Content

Patient-focused content receives strong citation rates because:

Comprehensiveness: Patient content covers questions from multiple angles (symptoms, treatments, side effects, lifestyle management)
Accessibility: Written in clear, non-technical language that AI models can easily parse and summarize
Regular updates: Patient websites frequently update content based on new research and patient questions
Question-based structure: Often organized around FAQs, which aligns with how users query AI models
Depth of information: Patient content tends to be thorough, addressing concerns that technical papers skip

Media Discussion Amplifies Patient Content Citations

The Ozempic case study revealed that broad media coverage significantly impacts citation patterns. When a pharmaceutical product receives extensive consumer media attention, patient-focused content citations increase dramatically, even for queries framed from a healthcare professional perspective.

This suggests AI models weight content based on:

Volume of available information on a topic
Recency and frequency of updates
Breadth of discussion across multiple sources
Engagement signals from readers and citations from other websites

When Regulatory and Official Sources Dominate Citations

Which Products Receive More Regulatory Citations?

Specialized prescription products without broad consumer markets see regulatory content dominate AI citations. The study identified clear patterns based on product type and market characteristics.

Dupixent Citation Breakdown (Biologic for atopic dermatitis/asthma):

Regulatory/Official sources: 50%+ of top-10 results across all DACH countries
Sources included: government health agencies, approval documentation, technical product information
Lower commercial competition due to specialist prescribing requirements
Limited consumer media coverage compared to Ozempic

Keytruda Citation Breakdown (Oncology – Austria):

IQWiG, EMA, G-BA, European Commission: approximately 60% of top-10 sources
Professional medical content: 19%
Patient information: 19%
Commercial content: minimal presence

What Determines Regulatory vs. Commercial Citation Balance?

The citation source mix depends on:

Commercial incentive: High consumer demand drives commercial content investment
Product accessibility: OTC or easily accessible products attract commercial publishers
Media coverage: Broad media discussion creates more content variety
Prescribing complexity: Specialist-only products see more regulatory/technical content
Market maturity: Newer products with limited available information default to official sources

Regulatory Content Citation Advantages

Official and regulatory sources maintain citation strength because:

Government health agencies carry inherent authority and trust signals
Technical documentation provides definitive product specifications
Regulatory approvals represent primary sources that other content references
Official sources rarely face SEO competition from commercial entities in specialized niches
AI models recognize regulatory domains as authoritative for specific query types

For niche products, specialized therapeutics, or B2B offerings, regulatory and technical content remains the dominant citation source across AI platforms.

Technical Accessibility Requirements for AI Citation

What Technical Factors Affect AI Citations?

Content receives citations only when AI systems can read, parse, and understand it. Technical barriers block even excellent content from AI visibility.

The study emphasized that pharmaceutical websites must address:

Site architecture: Clear navigation and logical content hierarchy
Content format: Text-based content rather than PDF-heavy pages
Access restrictions: Login walls and paywalls block AI crawler access
JavaScript rendering: Dynamic content generation can prevent AI crawlers from reading pages
Semantic markup: Structured data and clear HTML help AI models understand content context

How Do AI Crawlers Access Website Content?

AI platforms use specific crawlers to access web content:

OpenAI: GPTBot (training), OAI-Searchbot (search results), ChatGPT-User (user actions)
Perplexity: PerplexityBot (search), Perplexity-User (user actions)
Anthropic: ClaudeBot (training), Claude-Searchbot (search), Claude-User (user actions)

Websites must allow these crawlers in robots.txt files and ensure CDN/hosting providers don’t block them. Many websites inadvertently block AI crawlers through default security settings or outdated robots.txt configurations.

What Content Structure Do AI Models Prefer?

AI models cite content that features:

Clear section headings that indicate topic structure
FAQ formats that match conversational query patterns
Concise explanatory blocks rather than dense paragraphs
Lists and tables that present information systematically
Definition sections for technical terms
Primary source attribution and data citations

Technical documentation, academic papers with complex formatting, and marketing copy without clear structure receive fewer citations than well-organized, accessible content.

Frequently Asked Questions About AI Search Citations

What percentage of AI citations come from the top 10 websites?

Approximately 50% of AI citations come from the top 10 most-cited websites for any given topic. This concentration means the majority of visibility goes to a small number of authoritative sources. The remaining 50% of citations distribute across 40+ additional domains.

Do commercial websites get cited by ChatGPT and other AI models?

Yes, commercial websites receive frequent citations when they invest in SEO and content optimization. For products with high consumer demand like Ozempic, commercial providers (online pharmacies, telemedicine platforms) dominated citations, appearing in 3 of the top 10 most-cited sources.

How does traditional SEO affect AI citations?

Strong traditional SEO directly correlates with AI citation frequency. Websites that rank well in Google search results typically receive more AI citations because:

They have established domain authority
Their content is technically accessible to crawlers
They publish comprehensive, regularly updated information
They generate backlinks and engagement signals

Which AI platform cites the most diverse sources?

The study analyzed ChatGPT, Perplexity, Google AI Overviews, and Gemini. Citation patterns varied by platform, with some favoring regulatory sources and others citing more commercial content. Platform-specific optimization may be necessary for comprehensive AI search visibility.

Do patient information websites get cited for B2B or professional queries?

Yes, even for queries framed from healthcare professional perspectives, patient information websites received significant citations. For Biktarvy, 40% of top-10 results in Austria were patient-focused sources, demonstrating that AI models value comprehensive, accessible content regardless of the intended professional audience.

Can local websites compete with national domains in AI search?

Local websites can achieve top-10 placement in their home country but rarely cross borders. German domains dominated Austrian and Swiss results (40-60% of citations), while local Austrian and Swiss sites represented only 5-10% of citations. However, country-specific product sites (like Eliquis.ch) did achieve strong local placement.

What content types receive the most AI citations in healthcare?

Citation distribution varies by product and market:

High consumer demand products: 40-70% commercial/patient content
Specialized prescription products: 50-60% regulatory/official sources
Established therapeutic areas: balanced mix with professional publications prominent
Emerging treatments: higher percentage of academic and research citations

How often should content be updated to maintain AI citations?

Patient websites and commercial content providers that update frequently (monthly or more) maintained strong citation rates. Regulatory content remains cited based on authority rather than update frequency. The study suggests regular content updates correlate with sustained visibility.

Do AI models cite content behind paywalls or login requirements?

Generally no. Content behind login walls or paywalls blocks AI crawler access and receives minimal citations. Academic journals and technical documentation with restricted access showed significantly lower citation rates compared to open-access equivalents.

What is the relationship between media coverage and AI citations?

Broad media coverage significantly increases citation diversity and volume. Ozempic received extensive consumer media attention, which resulted in 70%+ of top citations coming from patient portals and commercial providers rather than traditional medical sources. Media momentum shapes AI citation patterns.

How to Apply These Findings Across Industries

AI Citation Patterns Apply Beyond Pharmaceuticals

While this study analyzed pharmaceutical products, the citation patterns apply across B2B SaaS, professional services, financial services, technology, and consumer brands. The core dynamics remain consistent:

Pattern 1: Source Concentration

The top 10 websites in any category capture 40-60% of citations
Breaking into top-tier citation status requires sustained investment
Domain authority, content depth, and technical optimization all contribute
Concentration exceeds traditional search engine result patterns

Application: Audit which websites currently dominate AI citations in your category. Identify gaps in their coverage that your content can fill. Build content specifically designed to address questions those top-tier sources don’t answer comprehensively.

Pattern 2: SEO Investment Pays Forward

Websites ranking well in traditional search receive more AI citations
Technical SEO fundamentals (crawlability, site structure, semantic markup) matter for both
Content optimized for search intent also serves AI model requirements
Strong backlink profiles correlate with citation frequency

Application: Continue traditional SEO work as the foundation for AI visibility. Websites without basic SEO hygiene won’t achieve AI citations regardless of content quality.

Pattern 3: Commercial Competition Intensifies

Where revenue potential exists, commercial content dominates
Multiple providers competing for keywords creates dense content ecosystems
Commercial sites often outperform academic or official sources in high-value categories
Investment in content marketing directly impacts citation share

Application: Expect commercial competition in any profitable category. Budget for content creation and optimization that matches or exceeds competitor investment. Monitor competitor citation rates using tools like OtterlyAI.

Pattern 4: User-Focused Content Wins

In pharma: patient content got cited frequently
In B2B: customer success stories, implementation guides, and user documentation perform well
In consumer: product comparison content, how-to guides, and review aggregation sites dominate
Content serving end-user needs outperforms pure marketing material

Application: Create comprehensive resources that answer customer questions thoroughly. Structure content around actual user queries rather than marketing messages. Include FAQs, troubleshooting guides, and implementation details.

Pattern 5: Geographic Citation Spillover

Dominant market content crosses borders into adjacent markets
Smaller markets see heavy influence from larger neighbors
Local content achieves strong placement within home markets only
Language and cultural context matter less than domain authority

Application: If you operate in multiple geographic markets, invest heavily in the largest market first. Content from that market will influence AI citations in adjacent regions. Build local content for top-tier placement in specific countries only when necessary.

Pattern 6: Technical Barriers Block Everyone

Fortune 500 companies with poor site architecture lose to startups with clean, accessible content
Login walls, paywalls, and restricted access block AI crawlers
PDF-heavy content receives fewer citations than web-accessible text
JavaScript-rendered content may be invisible to some AI crawlers

Application: Audit your technical infrastructure before creating new content. Fix crawler access issues, ensure robots.txt allows AI bots, verify content renders in text format, and eliminate unnecessary access restrictions.

Industry-Specific Applications

B2B SaaS Companies:

Official documentation and API references = regulatory/technical content equivalent
Customer community forums and user-generated content = patient content equivalent
Integration partners and technology blogs = commercial content equivalent Focus on comprehensive technical documentation, active community management, and partner content ecosystem development.

Professional Services:

Thought leadership content and industry research = regulatory content equivalent
Client case studies and implementation guides = patient content equivalent
Service provider directories and comparison sites = commercial content equivalent Publish original research, maintain detailed methodology pages, and create client success content.

Financial Services:

Regulatory filings and official disclosures = regulatory content
Financial education and customer resources = patient content equivalent
Comparison sites and review platforms = commercial content Ensure regulatory content is accessible, create extensive financial education libraries, and monitor review platform citations.

Technology Products:

Technical specifications and developer documentation = regulatory content
User guides and community forums = patient content equivalent
Review sites and tech news coverage = commercial content Maintain open-access technical documentation, support active user communities, and engage with tech media for coverage.

Specific Optimization Recommendations by Content Type

For Regulatory and Official Content:

Make all official documentation open-access and crawler-friendly
Use clear heading structures that match common queries
Include executive summaries that AI models can easily extract
Publish regular updates and clearly date all content
Structure technical information with defined terms and concepts

For Customer/Patient-Focused Content:

Organize content around frequently asked questions
Write in accessible language at appropriate reading levels
Update content based on actual customer questions and feedback
Include multiple content formats (text, tables, step-by-step guides)
Create comprehensive resource centers covering topics exhaustively

For Commercial and Marketing Content:

Invest in content marketing that provides genuine utility
Create comparison resources that fairly represent alternatives
Build detailed product information pages with specifications
Maintain active content publishing schedules
Ensure all marketing content is technically optimized for crawlers

Why AI Search Visibility Matters Now

The Shift from Click-Based to Answer-Based Search

The komm.passion study described this transition as “Shared Decision Making 3.0” in healthcare, but the pattern extends across all industries. Search behavior has fundamentally changed:

Traditional search model:

User enters query into Google
Clicks through to website
Reads content on original site
Returns to search results for more options

AI search model:

User asks question to ChatGPT, Perplexity, or AI Overview
Receives synthesized answer from multiple sources
May or may not visit source websites
Citation visibility determines brand authority

In this new model, citations replace clicks as the primary visibility metric. Users increasingly trust AI-synthesized answers over individual website visits. Brand presence in AI responses becomes more valuable than traffic generation.

Customer Behavior Has Already Changed

Your customers are asking AI about your products before visiting your website. Your competitors’ customers are asking AI where to find alternatives. Job candidates are asking AI which companies to apply to. Partners are asking AI which vendors to evaluate.

The pharmaceutical study demonstrated this: even healthcare professionals asking technical questions received answers influenced by patient content, commercial providers, and media coverage alongside regulatory sources. AI models synthesize across content types, meaning all your digital presence contributes to how you appear in AI responses.

The Competition for AI Visibility

This study analyzed five pharmaceutical products and found:

1,500+ unique citations across 150+ queries
50% concentration in top 10 sources per query
Significant variation by product, market, and query type
Commercial providers capturing 40-70% of citations in competitive categories

If pharmaceutical brands are competing this intensely for AI visibility, expect similar or greater competition in technology, financial services, professional services, and consumer categories. The brands investing in GEO now are establishing citation dominance that will be difficult for latecomers to overcome.

How to Measure Your AI Search Performance

Start With Citation Tracking

Before optimizing for AI search, understand your current performance:

Identify key queries: List the questions your customers ask about your product category, specific products, and brand
Track citation frequency: Measure how often your brand, website, or content appears in AI responses to those queries
Analyze citation context: Understand whether you’re cited as the primary source, mentioned alongside competitors, or absent entirely
Monitor competitors: Track which competitors receive citations and in what context
Measure across platforms: ChatGPT, Perplexity, Google AI Overviews, and Gemini may cite different sources

Key Metrics for AI Search Visibility

The study revealed these citation metrics matter:

Citation rate: Percentage of relevant queries where your brand appears
Citation ranking: Position of your citations relative to competitors (top-10 vs. top-50)
Source diversity: How many different pages from your domain receive citations
Citation context: Whether you’re positioned as authoritative, alternative, or cautionary
Platform coverage: Presence across multiple AI search platforms

Tools for AI Citation Monitoring

OtterlyAI tracks brand visibility across AI platforms and provides:

Citation frequency for your brand and competitors
Query analysis showing which prompts surface your content
Source attribution showing which of your pages get cited
Competitive benchmarking against category leaders
Platform-specific performance across ChatGPT, Perplexity, Google AI Overviews, and others

The pharmaceutical brands in the komm.passion | Team Farner study used OtterlyAI to identify citation patterns, measure competitive positioning, and track which content types achieved the strongest AI visibility. This measurement enabled them to optimize content strategy based on actual citation data rather than assumptions.

Get Started With AI Search Optimization

Track your current AI search performance before making optimization decisions. OtterlyAI shows you which prompts surface your content, which competitors outrank you, and where your citations come from across major AI platforms.

The pharmaceutical study analyzed 1,500+ citations to understand visibility patterns. You can start with a focused assessment of your brand’s performance on the queries that matter most to your business.

Sign up for OtterlyAI to measure your AI search visibility, track competitor citations, and identify optimization opportunities. The brands establishing citation authority now are setting the standard that others will struggle to match.