Fragrance shoppers are more curious than ever, with global fragrance sales hitting $40.4 billion in 2022, but how can consumers decide on the best bottle to purchase among the overwhelming number of bottles on shelves?
In previous years, scent entrepreneur Mindy Yang’s enterprise Perfumarie was created to be a neutral hunting ground. The NYC-based space presented shoppers with completely bare and unlabeled scent-testing stations so individuals could determine how they felt about a scent free from brand, packaging, or scent note persuasion. After all, how much does one’s subconscious or even conscious pre-shopping consumption of fragrance influencer reviews determine the bottle that lands in the bag? Puig recently launched a platform, WikiParfum, which offers customers brand-agnostic, machine-learning algorithm-driven perfume suggestions based on their favorite notes or past fragrances.
For the average in-store shopper, whether they visit a single-brand boutique or one stocking multiple shelves, the reality is there are only so many fragrance descriptions one can read and fragrance samples one can smell before fatigue sets in. According to a study by customer engagement platform Emarsys, 1 in 10 UK consumers break off a purchase transaction when overwhelmed by choice, with 24% defaulting to a previous brand purchase when feeling this overwhelm, limiting brand exploration. How to find the optimal fragrance in such surroundings? Kaorium by Scentmatic is hoping to offer them a solution.
An AI system that offers verbalization of fragrance qualities, the software is presented as a touchscreen with different unmarked bottles (20 available options for its larger model and 16 for its smaller size) for users to choose from. Simply going off of personal instinct, the program leads the individual down a vocabulary-driven scent exploration—utilizing broader terms like “romantic,” “unique,” and “energetic” rather than perfumery-only focused terminology—offering three fragrance recommendations at the end, complete with poetic descriptions of each.
While linguistic interpretation of scent can be highly individual, language is the prime medium of communicating scent, thereby helping shoppers tap into the type of fragrance they are looking for rather than being focused on a certain brand name or avoiding a certain material, for example. For retailers, the interactive interface offers an entertaining and different point of engagement for shoppers, especially those who might otherwise be too intimidated to consult a salesperson. Furthermore, Kaorium sees the tool as a communicator between the brand and consumer world, “respecting the cognitive diversity of the customer and providing an experience that allows the user to make a convincing choice by respecting the user's own feelings while providing the brand's definition of the product.”
To build the software, the Kaorium team collected data from consumers and professionals alike about the words that came to mind when thinking about fragrance, which were then put into their AI software. The machine-learning algorithm then selects the top 20 language expressions that are most likely to resonate with consumers, continuously expanding on and updating this vocabulary over time.
As Toshiharu Kurisu, CEO of Scentmatic, tells BeautyMatter: “We see three major systems of AI applications in the world: recommending fragrances that match the users’ tastes and preferences, AI that supports users' fragrance selection and educates their sensibility, and collecting data on users' sensory feedback and converting it into big data to create fragrances based on users' preference data.”
The invention was born out of a desire to update and customize fragrance shopping. “The experience of purchasing fragrances in physical stores has not changed over the past few decades. We believe that visuals, bottle design, staff advice, and brand story, are done in almost all stores and are generally not differentiated enough. The perception of fragrance is different for each person, and users do not always perceive fragrance as defined by the brand,” Shin Watanabe, Director at Scentmatic, adds.
During a proof-of-concept experiment at Tokyo perfume library Nose Shop, the company noted a 139% increase in store entries and a 287% boost in sales conversions. The system is already being employed in Shiseido’s R&D departments, proving the technology’s reach goes beyond the shopfront and can help create potential future blockbuster releases based on the user data it gathers.
It’s another interesting step into fragrance’s increasingly tech-driven future, be it on the side of creation or shopping assistance. The company also sees potential for the technology’s use in hair salons and at cosmetic counters to help consumers find the ideal products for their beauty needs. Kaorium is already being used across 10 stores in Japan; will its entry into European and US markets be next?
