Before the project began, there was no individualized approach across all channels - neither in the web store, nor in stores or newsletters. Customer data from the website, store and CRM was not being used, meaning that valuable information about customer behavior and needs remained unused.

Especially during the COVID-19 pandemic, when stores had to remain closed, the sales potential was not fully exploited due to a lack of personalization.

‍

‍

We implemented a GCP-based Customer Data Platform (CDP), linked to CRM systems and the Magento online store via automated pipelines. Customer interactions were systematically recorded on the basis of a newly created data layer in the web store. This resulted in a 360° customer profile that integrates online and offline behavior.

A recommender system based on collaborative filtering uses this data to generalize product recommendations to similar customers. The machine learning pipeline is continuously supplied with new data (e.g. clicks, purchases, shopping cart, try-on, carousel, reservations...) and automatically retrained. The recommendations are integrated into the webshop, CRM and newsletter system via an API - so all channels receive up-to-date, personalized suggestions.

‍

Customer data is now actively used for personalized recommendations - with a noticeable impact on relevance and sales.