Dev.to
7/14/2026
Study of 2,000 LLM query-results: new brands cited 0%, Wikipedia presence is top citation predictor
Original: I ran 500 brand queries across ChatGPT/Claude/Gemini/Perplexity. New brands get cited 0% of the time. Full data inside.
Short summary
Original research across 2,000 query-results on ChatGPT, Claude, Gemini, and Perplexity finds brands under one year old are cited 0% of the time. The strongest predictor of citation is a Wikipedia/Wikidata entry, while llms.txt and JSON-LD schema had negligible impact. High-DA offsite mentions across 3+ domains mattered more than any on-site factor.
- •Brands <1 year old cited 0% across all four LLM engines
- •Wikipedia/Wikidata entry is the strongest citation predictor
- •llms.txt had zero measurable effect; offsite high-DA mentions mattered most
Generated with AI, which can make mistakes.
Is this a good recommendation for you?
