Search papers, labs, and topics across Lattice.
This study analyzes how large language models (LLMs) source brand reputation by examining 167,551 URL-grounded citations across 128 brands, 12 home markets, and 13 languages. The findings reveal that LLMs predominantly rely on third-party sources, with 85.7% of citations coming from non-owned domains, and highlight the concentration of citations, where 80% originate from just 18% of domains. Notably, Wikipedia emerges as the most-cited source in 11 out of 12 languages, underscoring the significant influence of specific platforms on brand visibility in AI-generated responses.
LLMs overwhelmingly rely on third-party sources for brand information, with Wikipedia dominating citations across most languages.
When a large language model (LLM) answers a question about a company, it grounds the answer in retrieved web sources, and those sources decide what the model says. Most analysis of AI brand visibility looks at the answer text. This study looks one step earlier, at the citations. We merge three Rankfor.AI datasets covering 128 brands across 12 home markets and 13 languages, and analyse 167,551 URL-grounded citations (189,974 total attribution rows). We classify each citation by domain and source type and measure where AI gets its brand information, by language and by market. Four patterns hold. First, AI grounds brand answers overwhelmingly in third-party sources: 85.7% of citations point to sites the brand does not own, against 14.3% owned. Second, the source base is concentrated and long-tailed: 80% of citations come from about 18% of domains, fitting a Zipf law (alpha = 0.86, R^2 = 0.983). Third, one reference site dominates almost everywhere: Wikipedia is the most-cited domain in 11 of 12 languages, the exception being Lithuanian, where the business daily vz.lt edges it (4.38%). Fourth, the source mix is market-specific at the margin: for 46 Polish national brands the most-cited domain is YouTube, and four HR and careers portals supply 637 citations against 297 for Polish Wikipedia, about twice as many.
