Case Study: Optimising Marketing Spend using Predictive Data Analytics

Predicting Customer Response using Data Analytics

17% saving of annual catalogue mailing costs, through improved targeting

A luxury fashion retailer optimised their marketing spend by using a machine learning model that predicts who is most likely to respond to catalogue sales campaigns with high accuracy.

How did we help?

Using data describing their customers’ purchasing habits, we were able to deliver a self-service machine learning solution which this company can use in advance of each marketing campaign to predict who is likely to make a purchase, thus allowing them to improve their targeting, save costs and smartly reallocate their marketing budget towards other initiatives.

The result

  • A highly accurate model resulting in saving 17% of annual catalogue mailing costs while losing fewer than 0.5% of buyers
  • A range of options to tune how the model makes predictions to fit their needs
  • Additionally we advised on uncovered data discrepancies, contributing to a process improvement programme to ensure that their data will be captured consistently and accurately

“Our catalogue marketing campaigns are hugely important for our business. With the model that Ecovis delivered, we are now able to select our catalogue audience according to their likelihood of responding. We expect this to return to us an annual net saving of approximately 17% of our annual catalogue mailing costs for minimal loss of demand. The data analytics team at Ecovis understood the scope of the work, were professional and patient, and delivered exactly the end product that we needed. This was a dream project for us, and I look forward to having their ongoing support.”

CEO
Luxury Fashion Retailer