Read how GSPANN helped a US-based skincare products manufacturer with over 2 million customers and 1.5B annual sales in developing a machine learning (ML) model to predict customer churn. Our advanced analytics team created the model utilizing predicting capabilities of its in-house Keras-based ML solution called ‘Zeolite.’ Key highlights: The solution helps the client in running personalized campaigns to retain the preferred customers who form a major part of their sales. These campaigns are aimed at keeping these customers engaged and sales network active.  When the client sent personalized offers to the customers who were selected through the ML model, sales made from them were 82% higher and the average revenue per customer was 80% higher compared to the customers who weren’t sent any offers.  Personalized campaigns on the preferred customers helped in increasing the average order value by 9% and the purchase rate by 60%. This case study can help you understand how an ML model can help you predict customer churn and take necessary actions to prevent losing them, and thereby, any revenue loss. Developing such a capability can prove to be a boon for any e-commerce business.