154,944 AI Citations Reveal a Fragmented Landscape for B2B Brands
The index finds significant variation in AI-generated answers, highlighting the importance of authoritative sources and continuous measurement of brand visibility.
AI-generated answers to the same B2B executive questions changed substantially 50% to 60% of the time, according to PAN Communications‘ C-Suite Signals: AI Citation Index.
The study analysed 31,500 responses collected from ChatGPT, Claude, Gemini and Perplexity between late August and late September 2026, examining how AI platforms respond to questions commonly asked by C-suite executives.
The findings highlight the challenges brands face in maintaining consistent visibility across AI-generated responses. A brand appearing in an AI answer at one point does not guarantee that it will remain visible in subsequent responses.
For marketing and communications teams, the research suggests that measuring AI visibility requires ongoing monitoring rather than one-time assessments, alongside greater investment in credible third-party sources.
“Leaders want to know whether their brand shows up when an executive asks AI for advice, and our data shows that answer can change from one week to the next,” said Mark Nardone, CMO of PAN Communications.
“The brands that hold their place are cited by sources AI already relies on, like analysts, trade outlets, and professional associations.”
Authoritative Sources Show Greater Citation Consistency
The report found that 42% of sources cited during the first week remained in AI-generated answers by the fourth week, although retention varied considerably depending on source authority.
The National Institutes of Health (NIH) retained 76% of its citation positions, while Deloitte and the Society for Human Resource Management (SHRM) maintained 68% and 67%, respectively.
In comparison, community-driven platforms recorded substantially lower retention rates, with Medium at 27% and Reddit at 21%.
These differences suggest that established institutional and professional sources may offer brands more sustained visibility within AI-generated responses.
The research also identified significant variation across executive roles. Citation variance averaged 54% across nine C-suite personas, ranging from 39% for Chief Human Resources Officers (CHROs) to 70% for Chief Medical Officers (CMOs).
Among the four AI platforms examined, ChatGPT delivered the most consistent answers, with lower variation than Claude, Gemini and Perplexity.
For marketers, the findings reinforce the importance of building brand authority through sources that AI systems repeatedly reference, rather than relying solely on content volume.
AI Citations Span Thousands of Sources
The study identified 154,944 citations across 16,503 unique domains, illustrating the breadth of information sources influencing AI-generated responses.
LinkedIn emerged as the most frequently cited domain, although it accounted for just 5% of total citations.
The report also found that citation patterns differed according to executive responsibilities. For Chief Medical Officers, NIH and PubMed collectively accounted for 23% of citations. CHROs frequently encountered SHRM, while CIOs and CTOs relied more heavily on CIO.com. Gartner was a prominent source for CMOs and Chief Revenue Officers.
These differences indicate that AI visibility strategies may need to account for the specific information sources trusted within each executive function.
Rather than pursuing visibility across every available channel, brands could benefit from identifying the publications, professional associations and industry resources most relevant to their target audiences.
AI Visibility Gains Extend Beyond Brand-Owned Content
PAN Communications also highlighted client programmes demonstrating how earned media and thought leadership can contribute to AI visibility.
For payments technology company Nuvei, the agency combined a Mexico market launch with original consumer research and thought leadership involving seven executives.
Nuvei’s AI Visibility Score increased by 4.8 points to 14.4%, while earned media coverage generated 59,598 AI citations, nearly three times the number attributed to its own website.
A separate programme for DynaMed combined paid, earned, shared and owned media with executive LinkedIn activity.
The campaign delivered a 1,500% quarter-over-quarter increase in organic keyword rankings, more than 450 page-one and AI Overview keyword placements, and a 60% rise in organic search sessions.
For Tungsten Automation, PAN developed a US and UK media programme focused on enterprise AI adoption, securing coverage across publications including CIO.com, InformationWeek, Quartz and TechTarget.
The examples illustrate how integrated communications strategies can support brand discoverability across traditional search and AI-generated answers.
Key Findings From the AI Citation Index
- AI answers frequently change: Responses to identical executive questions varied substantially 50% to 60% of the time across the study period.
- Authoritative sources retain visibility: NIH maintained 76% of its citation positions, compared with 27% for Medium and 21% for Reddit.
- Citation patterns vary by executive role: Average variance reached 54% across nine personas, with Chief Medical Officers experiencing the highest variation at 70%.
- ChatGPT delivers greater consistency: Among the four platforms analysed, ChatGPT showed the lowest variation in its responses.
- AI draws from a broad source ecosystem: The research recorded 154,944 citations across 16,503 domains, with LinkedIn accounting for 5% of the total.
The findings suggest that AI visibility is becoming an increasingly important consideration in B2B brand and communications strategies.
As AI platforms influence how executives discover information and evaluate potential partners, marketers may need to measure not only whether their brands appear in generated answers, but also how consistently they are referenced and which sources contribute to that visibility.
The report points towards a broader shift in brand measurement, where sustained credibility across authoritative sources becomes as important as visibility at any single point in time.
ALSO READ: The Trust Gap in Personalisation
