Big Data

Predictive Modeling To Allocate Marketing Spending

Download Now Date Added: Jun 2010
Format: HTML

A student education and tutoring company wanted to understand how well different media channels attracted students. What was the cost to acquire a new student through the Web? Which types of media were most efficient? Did spending on traditional media drive Web traffic? The company believed that better allocating marketing expenditure could help it attract leads and drive enrollment more efficiently and lower costs, while protecting enrollment levels. The company's learning centers offer personalized instruction in reading, writing, math, study skills and test preparation. It markets its services through various channels, including television, radio and Web. The company wanted to better understand how it was spending its marketing dollars and the relationship between the cost to acquire a student and the economic value of that enrollment. It also wanted to quantify which channels (phone calls, Web-generated calls and Web form hits) are most effective at generating prospective students, both under its current marketing plan and under a new plan that allocated marketing costs and forecast future contacts with the help of analytical modeling. Deloitte practitioners began by helping the company compile historical data on its marketing expenditures and response patterns across the various channels, including looking at response rates when the company used multiple media channels simultaneously. Using exploratory data analysis and regression modeling techniques, we were able to help the company quantify the relative effectiveness of spending for each channel as well as the benefits of cross-channel spending, such as instances where radio advertising led to an increase in Web hits.