AI model drives 40% product recommendation uptake for COTY
Challenge
Luxury fragrance company Coty wanted to transform the online experience for its customers. Their aim was to make buying fragrances as intuitive and engaging as buying clothes.
To do this they wanted to create a recommendation engine which would automatically serve customers with perfumes and aftershaves perfectly tailored to their tastes.
However, due to the very personal and often intangible nature of fragrance preferences there was no ‘out of the box’ solution to make this happen.
Solution
Recognising that such a complex recommendation engine had never been built for the fragrance industry before, Profusion set about creating a strategy which would leverage ‘traditional’ sources of consumer data and more in depth personal information as the basis of training its AI.
To do this Profusion conducted an extensive 200-question survey of luxury fragrance buyers, to understand which psychometric and personality profiles aligned with fragrances preferences.
A team of data scientists at Profusion then built a series of five machine learning models to help create a fully personalised customer product recommendation engine.
Result
The tool was first adopted by in-store sales associates – 4 out of 5 users rated their top recommendation at 4 stars or higher.
40% went on to purchase that recommended product.
Sound good?