Some styles and trends are easy to explain—maybe you want 1990s-inspired high-waisted jeans paired with a classic oxford shirt for example. But many more of the quickly shifting fashions and trends, whether in clothing, home décor, or other domains, really are more of a “vibe.” Consumers know what they mean, and other people likely can infer what they want, but it’s still hard to describe styles like “cottagecore” or “alt formal” precisely.

Such lack of precision is a barrier for most technology- and AI-enabled recommender algorithms maintained by retailers. If users type “VSCO preppy” into most retailers’ search pages, they are unlikely to uncover links to the brightly colored animal prints they seek or the Roller Rabbit brand that they aspire to own. For this kind of service, shoppers would still need human store clerks to help them define and find their style.

But a new technology, called Shopping Muse, aims to make the inspiration searchable, as long as it is installed on the retailer’s site. If a visitor to that site notes their interest in “kingcore” or “queencore” for example, Shopping Muse can leverage additional data about the shopper (e.g., demographics, prior purchases) to identify just the right cloak and sash that the consumer is likely to appreciate. In addition, Shopping Muse will provide a link to the furniture department of the retail site, highlighting the throne-inspired dining room furniture it lists, as well as the jewelry page, with its selection of ornate, regal-looking rings and bracelets.

The searches are not limited to descriptions of specific aesthetics though. Shopping Muse’s own website offers a multistep search as an example: Imagine a user requests help finding an outfit for a wedding in Bali, held during this upcoming winter. The request produces a result page with a set of about four possibilities. In response, the user asks for edgier options, “like a pop star would wear.” Then among the consideration set, the user can specify, “Show me more like the black option,” which brings up a wider selection of outfits that reflect a bold, wedding-appropriate style for a colder weather event in a tropical climate. But in all these steps, the user’s language and requests are informal and casual, similar to what they would use when chatting with a live fashion expert, rather than requiring the specific terminology demanded by many search engines.

Shopping Muse is a product of a company called Dynamic Yield, which itself recently was acquired by Mastercard. By offering this AI-enabled personalization service to its business clients, the credit service provider might encourage them to push the use of Mastercard accounts to buyers. Furthermore, it might help its consumer clients shop more easily (and use their Mastercards) to make purchases that more accurately reflect their preferences and desires.

Discussion Questions

1. Do you have a personal aesthetic? How would you describe it? Would a search engine that could use your informal description to produce product recommendations be helpful to you?
2. How should Mastercard charge retailer adoptees to install this AI tool on their websites?

Sources: Tatiana Walk-Morris, “Mastercard Launches Shopping Muse, an AI Tool that Gives Product Recommendations,” Retail Dive, December 12, 2023; “Dynamic Yield by Mastercard Unveils Shopping Muse, the Next Generation Personal Shopping Assistant,” press release, November 30, 2023, https://www.mastercard.com/news/press/2023/november/dynamic-yield-by-mastercard-unveils-shopping-muse-the-next-generation-personal-retail-assistant/; Dynamic Yield by MasterCard, https://www.dynamicyield.com/shopping-muse/