A new study by Human Forward has found that 48% of assessed cycling brand websites meet its criteria for AI shopping readiness, while 55 of the 168 brands tested received no recommendations from any of the eight AI assistants surveyed.
Published in November 2026, the research examined 168 cycling brands operating in the UK, comprising 69 British brands and 99 international brands. The study covered bicycle retailers, manufacturers, e-bikes, clothing, components, accessories and sports nutrition.
Of the 147 websites that Human Forward could assess, 70 had either a Universal Commerce Protocol (UCP) manifest or a storefront Model Context Protocol (MCP) endpoint, the standards used in the study to identify websites ready for AI shopping agents. This represented 48% of assessable websites, compared with 16% in the company's separate study of 200 major US retailers and brands.
Cycling clothing recorded the highest readiness rate, with 18 of 25 assessed websites meeting the criteria. Accessories followed at 13 of 20, while bicycle manufacturers recorded 19 of 55. Among retailers, five of the 15 assessed websites qualified.
The study also found significant gaps in brand recognition. When asked to identify SRAM, all eight AI assistants described it as static random-access memory rather than a bicycle component manufacturer. However, all eight recommended SRAM when asked to name leading cycling component brands.
Overall, 33 of the 168 brands were not correctly identified by any of the assistants when queried by name alone. Meanwhile, 55 brands received no recommendations when the assistants were asked to identify the best brands in their respective cycling categories. These included Colnago, Scott, Orbea, Ridley and Café du Cycliste.
The research also examined brand-owned rider communities. Of the 152 homepages that Human Forward could read, only 12 linked to rides, events or member communities operated by the brands themselves. Only five of those 12 brands ran their own member communities. In comparison, 35 brand websites linked to Strava groups.
The study was conducted on 6 November 2026, using ChatGPT, Claude, Perplexity, Google Gemini, Grok, DeepSeek, Meta Muse and Mistral Le Chat. Five questions covering brand identity, where to buy, category recommendations, brand legitimacy and product ranges were submitted to the assistants.
Human Forward founder Peter Abidi said cycling brands were better prepared for AI shoppers than much of the wider retail sector, with smaller brands leading the way. However, he identified brand recognition and brand-owned communities as areas where further progress is needed.
The report highlights that being technically accessible to AI shopping agents does not necessarily mean a brand will be correctly identified or recommended by AI assistants.
