Predictive Analytics Furniture

Furniture Retailer Revamps Mover Marketing Program Using a Response Model


Download This Case Study



A regional furniture retailer had been supporting a direct mail program targeting both current and prospective customers who were preparing to move or had recently relocated. The campaign targeted a select group of customers based on household income and home value. The retailer wished to acquire more customers, generate additional revenue per customer, and improve overall ROI.

The Process

The company engaged Speedeon Data’s predictive analytics team to develop a response modeling featuring new movers. The model incorporated geodemographic variables at the ZIP+4® level[1], including:

  • Wealth demographics
  • Urbanicity and location type
  • General demographic attributes, such as: age and level of education
  • Size of home

A machine learning algorithm was then applied to the model to help identify consumers most likely to respond.

The Result

The response model was able to predict likelihood of response at various mail volumes. The model enabled the furniture retailer to accurately forecast new customer growth along with associated revenue and marketing costs at different levels of customer acquisition.

Overall campaign effectiveness improved by 300 percent as calculated by a function of program cost to generated revenue. The retailer continues to see increased sales and customer growth as it uses the response model in its mover campaigns.


[1] ZIP+4® level is a registered trademark of the United States Postal Service