Brand research
The Shape of LLM Awareness
LLMtel compared 86,190 answers from 17 models to 5,070 prompts about 1,000 entities. See how differently the models distributed their attention.
When Wikipedia Fails: Three “Known‑but‑Never‑Named” Case Studies (From a 17‑LLM Benchmark)
Three cases in a 17-model benchmark show why appearing on Wikipedia does not guarantee a mention in an AI answer.
The Wikipedia Multiplier (Inside the “Ignored” Bucket): Why Wikipedia Still Separates the “Known” From the “Barely Known”
Among brands models recognized but rarely mentioned, a Wikipedia entry was associated with stronger recognition. Explore the results of the study.
The One Top‑25 Entity That Didn’t Need Wikipedia
One entity reached the top 25 in LLMtel’s benchmark without a Wikipedia page. See which other evidence helped models identify it.
The Ignored: Why LLMs Know You But Won’t Say Your Name
Across 86,190 answers, LLMtel found brands that models could recognize by name but seldom brought up on their own.
The Cliff: Why Most Brands Barely Show Up in AI Answers
In LLMtel’s benchmark, a few brands captured many mentions. See where the steepest drop occurred among the remaining brands.
Silent Overachievers & Quiet Giants
A benchmark of 1,000 entities across 17 models reveals two surprises: recognizable brands that rarely get mentioned and less familiar names that do.
From “Ignored” to “Named”: What Makes AI Actually Mention an Entity?
Some entities in LLMtel’s 1,000-entity study were recognized when named directly and overlooked in open-ended answers. What distinguished the ones that earned mentions?
AI Doesn’t Just Read Wikipedia, Wikipedia Stabilizes Your Name
A chatbot can treat two spellings of a company name as different entities. See how Wikipedia can help it connect those names.
