AI-Generated Fragrance Discovery

Currently at two locations in Harajuku, Tokyo, the blind smelling experience focuses on unique preferences, enabling visitors to intuitively discover scents and better express the type of fragrances they like. The user chooses four perfumes from 16 bottles, and the AI displays a host of descriptors – like “dynamic” and “enchanting” – to convey the impression of each selection. As the exercise is repeated, the person can choose words they feel best correlate with the scents they picked, allowing the AI to analyse their favourites and generate a personal poetic phrase to express their ideal aroma. For example, when Stylus went, our profile was defined as “flag fluttering the blue sky”.

The user is then given a QR code to access a range of recommended perfumes, as well as related words they can use to describe their preferred fragrance. Scentmatic’s success is evidenced by positive in-store sales, with an increased conversion rate of 287% after customers used the Kaorium system. It’s now seeking retail partners in London.

The technology’s capabilities stretch further than a discovery experience, by providing data analysis that can be used for insights and product development. With every consumer interaction, the AI’s vocabulary continues to expand, and this can be used for marketing and perfume innovation. Additionally, Kaorium is not limited to the fragrance field, with the potential to be used across multiple industries, including personal care, food and beverage, and education.

Brands like Scentmatic are paving the way for future scent development that will instil moments of delight, while ensuring individuality and personal preferences are at the forefront of experimental multisensory experiences. For more tech-lead fragrance innovations, see The Brief.