AI Models Fail to Cite Emerging Brands
A comprehensive study of 500 brand queries across leading AI models—ChatGPT, Claude, Gemini, and Perplexity—reveals a critical blind spot: emerging brands are completely invisible to these systems. Researchers at Kre8on, a platform that monitors brand mentions in AI responses, found that brands less than one year old were cited 0% of the time across all tested engines. This suggests a significant challenge for new entrants trying to gain visibility and establish credibility in an AI-driven information landscape.
The research methodology involved 50 distinct brands, categorized by age: 0-1 year, 1-3 years, 3-5 years, and 5+ years. For each brand, 10 different intent queries were executed, ranging from general inquiries like "best [category]" to specific comparisons such as "[brand] vs [competitor]". This resulted in a total of 2,000 query-result pairs analyzed across the four AI platforms. The primary metric logged was the citation rate, specifically the source domain from which the AI drew its information.
Key Predictors of AI Citation
The study identified the single strongest predictor of whether a brand would be cited by an AI model: the presence of a Wikipedia or Wikidata entry. This indicates that structured, authoritative knowledge bases play a disproportionately large role in shaping AI's perception of brand relevance and trustworthiness. Brands with these established digital footprints are far more likely to be recognized and referenced.
Interestingly, the research found that traditional on-page SEO signals had a comparatively minor impact. Factors like the presence of an `llms.txt` file or the implementation of JSON-LD schema resulted in zero measurable effect on citation rates. This is a surprising detail; many SEO professionals assume these technical elements are paramount for discoverability. Instead, the study found that high Domain Authority (DA) off-site mentions across multiple domains (three or more) were more influential than any on-site optimization efforts. This suggests a pivot towards off-site authority and established digital presence over technical SEO for AI visibility.
Implications for New and Established Brands
For brands under a year old, the findings are stark. Without a Wikipedia presence or significant off-site mentions across reputable domains, they simply do not register in the current AI information ecosystem. This creates a difficult feedback loop: new brands struggle to gain the off-site mentions and authority needed to be cited by AIs, which in turn makes it harder for them to gain visibility and acquire new customers through AI-powered search and discovery.
Established brands, particularly those with existing Wikipedia pages and a strong network of high-DA backlinks, are positioned to benefit significantly. They are already recognized by the AI models and are more likely to be included in relevant search results and summaries. This creates a moat around established players, potentially making it harder for new competitors to break through the noise. The reliance on Wikipedia as a de facto arbiter of relevance is a critical insight for any brand aiming to improve its AI discoverability.
Methodological Rigor and Future Research
The researchers were transparent about their methodology, encouraging scrutiny and further investigation. The breakdown of brands by age and the variety of intent queries ensure a robust dataset. The analysis across four major AI engines provides a broad view of the current state of AI citation practices. However, the study raises further questions about the long-term implications of this bias. What happens to innovation and market dynamism when AI systems systematically favor incumbents? Will AI developers actively work to diversify their citation sources, or will this bias become entrenched?
This research underscores the need for new strategies for brand building in the AI era. Simply focusing on traditional SEO may not be sufficient. Brands need to prioritize building genuine off-site authority and, where possible, aim for inclusion in structured knowledge bases like Wikipedia. The current AI information architecture appears to be a closed garden for new entrants, and understanding the keys to entry—established authority and structured data—is paramount.