AI tools are becoming common in perfume design and testing

In short: AI is increasingly being used to create personalized perfumes and to help fragrance companies work faster and cheaper.

What's going on

Some perfume brands now use AI to help customers create a custom scent. EveryHuman, a brand founded in 2018, asks shoppers an online set of questions and then matches the answers to a set of scent “building blocks” to mix three small custom perfumes. The company also has in-store machines in the UK that blend the perfumes on the spot, like a coffee machine that makes different drinks from the same ingredients.

Another company, Scircle, launched in 2025 and uses an AI chat to ask what you want your perfume to smell like. The founder says people can ask for very personal scents, even something as specific as a memory. A 20ml bottle costs $95, and it is sold online and in a few physical locations.

Behind the scenes, AI is also being used to speed up product development and reduce costs. Moodify says it has digitized about 2,800 perfume ingredients and built a model that predicts how combinations will smell (a bit like a recipe calculator). The company says it can help brands reformulate perfumes to meet new rules and cut costs by using lower-cost ingredients, and it claims it can shrink a new scent timeline from about 18 months to three.

Other firms focus on predicting whether people will like a fragrance before it launches. FoodPairing AI says it created “digital twins” of consumer types, meaning software-based stand-ins that react like different kinds of shoppers, with results it says match real consumer feedback about 77 percent of the time.

What to watch

Big ingredient suppliers like Givaudan have built their own internal tools, but they say AI is meant to assist perfumers, not replace them. The next question is how far “digital smell” can go, and whether these tools will change what perfumes cost and how quickly new ones appear.

Source: Financial Times