How Kaorium is disrupting the fragrance industry with AI
By Kaorium
Published · 3 min read

Fragrance writer, Nicola Thomis, sat down with Ben Janoušek, KAORIUM's CEO to find out how this new technology aims to change the landscape for fragrance shopping.
Nicola: A lot of the emerging uses of AI in the fragrance industry have centred around scent creation. How is KAORIUM different?
Ben: KAORIUM is different because it focuses on the person buying fragrance, rather than the creation of the product itself. We make it easier for people to cut through the noise of a busy fragrance market, and find scents they love in a fun, interactive way.
N: Tell us more about the journey that KAORIUM takes people on?
B: KAORIUM is unique in the way that it presents scents to the user. Before anything is smelled the customer picks a phrase which matches the mood or style they are trying to achieve. KAORIUM then suggests a group of three to four fragrances for them to smell. The customer then picks words or moods they associate with those fragrances. For example, a user might smell a woody fragrance, and they might associate positive words with that such as “rich” or “deep”. This process is repeated several times with selected scents chosen by KAORIUM based on the customer responses. The AI going on in the background uses thousands of data points to guide the user towards scents that they might enjoy based on the information they have put into the system.

N: What challenges does KAORIUM address?
B: Scent is really difficult to describe for people and there is no one set of correct answers to describe how any particular scent smells. What I might describe as a “lemon” fragrance, someone else might describe as “happy” or “energetic”. The first challenge that KAORIUM addresses is getting users to think about and engage with the language that they associate with particular scents.
The second challenge is working out what consensus there is from fragrance experts and enthusiasts as to how they describe fragrance, and where that matches or differs from how users describe it. KAORIUM then takes all these different sources of information and finds the commonalities. This makes it a lot easier for the user to then communicate in a way that feels natural to them, but helps them tap into this shared language about a particular scent.
Whilst it is doing this, KAORIUM is also learning what the user likes and dislikes, ultimately guiding them towards scents containing those preferences. For example, people might ask for scents that are “milky”, but they might not know that sandalwood, that is listed on the notes, can give that effect. KAORIUM removes the need for them to make that link, and instead takes them straight to the fragrance containing the feel that they are looking for.

N: What’s the benefit of this over a more traditional shop experience though?
B: In the KAORIUM process, scents are all smelled blind, so users are reacting to just the fragrances and the word prompts. They aren’t bombarded with notes lists, fancy bottles or flashy advertising. This means that they are more likely to find a scent that they genuinely love and which will continue to satisfy them once they’ve left the store.
For stores it’s a useful device too. KAORIUM is a way for sales assistants to get inside the mind of their customers, helping them start conversations about the scents and allowing them to guide customers successfully. KAORIUM also helps quickly narrow down what the customer is looking for without the nose getting too fatigued or them going “nose-blind” - something we know that it is important to avoid to have a happy and productive in-store experience.
About the editorial team
Written by
Kaorium
Articles from the Kaorium team, the people bringing the power of AI to fragrance consultation.


