EN   FR
claraVista



Case Study
—

Enabling customer personalization through machine learning

Product recommendations, next purchase date, replenishment date…

←   DATA & AI STRATEGY

Challenge

1 —

A well-known French cosmetics brand

2 —

The desire to personalize the customer experience to better meet their needs

3 —

Numerous possible use cases for personalization

The Key Questions

1 —

How to predict individual customer needs?

2 —

How to make the various scores available to business teams according to each use case needs?

3 —

How to adapt each score to its intended use?

Approach

1 —

Machine learning modeling using data from various sources of the brand to best predict each individual score and comparison of numerous models

2 —

Deploying the scores for regular updates and direct use by business tools (website, campaign management tool)

3 —

Monitoring performance with systematic A/B testing

Results

1 —

A carousel of personalized recommendations on the brand's websites

2 —

Des campagnes emails plus personnalisées, envoyées à chacun au bon moment et avec les bons produits

At the heart of the subject

Beyond pure machine learning modeling, it was important to co-create each of the scores with the business teams that would use them, ensuring a perfect fit for their needs and that the scores would be used effectively and provide value. For these same reasons, questions about performance monitoring and the availability of the produced scores have also been central.

Case Study :
Define the roadmap and the data architecture of a luxury brand

Case Study :
Set up a platform for data sharing/valorization