Nikkei CrossTrend's AICAS survey asks: what sells via AI?

Nikkei CrossTrend put forward AICAS in April 2026 as a way to describe shopping that begins with a question to a generative AI rather than a keyword search. Its premise is that the transition is not even, and that some products will be recommended by AI while others will not.

To test that, the publication paired with Macromill on a survey of 1,000 people spanning 44 product categories. The early case installments already lean toward categories where AI answers carry weight: Gurunavi's restaurant app UMAME! for choosing among 540,000 restaurants, Yahoo! Shopping's push into apparel, HIS on in-store close rates, Hotel Monterey on ChatGPT recommendation rates, and Rakuten Travel on review volume.

A follow-up installment framing the divide as four points of "problem solving" points to the practical question for marketers: whether their product surfaces in an AI reply at all, and, if it does, what it takes to keep it there. Whether that pattern holds at the category level should become clearer as the survey's category-by-category data is released.