Audit of 100 businesses finds big national brands lag indie firms in AI recommendations

A controlled audit of 100 businesses conducted on May 31, 2026, scored how well their websites provide AI systems like ChatGPT and Claude with the information needed to recommend them. The study spanned local, national, international, and niche B2B/SaaS companies, and found that national brands averaged the lowest score at 75.5, while international businesses led at 91.5. Small businesses such as Bold Street Coffee achieved perfect scores, while major UK brands like Pret A Manger and Greggs scored poorly, largely because their JavaScript-heavy sites make it difficult for AI models to extract factual content. The most common shortcomings across all sites included missing FAQ sections, reviews not present as readable on-page text, and unclear pricing information. The author notes the findings reflect a model of what AI systems reward rather than confirmed data from model providers, but argues that as AI-generated answers increasingly displace traditional search results, visibility to AI recommenders is becoming a meaningful business concern.