One of our clients launched a co-branded rewards credit card with a national bank, resulting in millions of new members. After a few years in place, the client wanted to use more sophisticated marketing tools to generate increased value from the program. The immediate opportunity was to move away from a “one size fits all” marketing approach toward a differentiated, segment-level approach.
The firm’s most often used marketing lever (and biggest investment) was price discounting, offered in the form of “bonus rewards” to its co-branded credit card customers. The challenge was to determine how much incentive to offer an individual customer for different products and how to optimize program performance against certain constraints such as budget limitations. In addition, the client wanted a flexible tool that would help it structure campaigns to meet corporate objectives, which might alternate from maximizing sales volume to maximizing profitability.
We created a solution that combined results from prior marketing campaigns with new behavioral models in a mathematical optimization framework. The solution uses segment-specific statistical models to predict the likelihood of a purchase based on different marketing incentive offers in each segment. The client can now identify the correct incentive amount to offer card members in each segment, in order to reach the maximum achievable sales or profit for a campaign. The system literally considers millions of possible scenarios of offer levels and finds the best solution. And each campaign can now be optimized for different goals, including total sales volume, incremental sales volume, profitability, and marketing expense.
The tool quantifies the cost-volume tradeoffs associated with multiple offer structure scenarios. Through use of the tool, the firm can confidently design programs that will maximize sales, profit, or both. Since the launch of the solution, campaign performance has been lifted significantly, and the accuracy of result forecasts has improved markedly.