Booking a Cross-Sell
Regardless of the cross-sell product, the front-end channel would initiate a cross-sell offer action through CRM. An offer action included electronic mailers specifying the loan or credit card amounts being sent to customers through the centralized marketing team and outbound telephone calls from the call center.
The company relied heavily on Dunia’s outbound telesales team to contact eligible customers and ensure cross-sell would happen. While the outbound team ensured that call-out would happen, the marketing and product team ensured that customer communications were sent out via all channels (such as e-mail, text message, and direct mail) for the broadest reach possible. The process worked seamlessly across all units—SAU, credit, outbound calling, marketing, and operations. The analytics team played a pivotal role in ensuring well-orchestrated coordination across all involved units. This called for significant soft skills over and above the technical skills expected of analytics.
Customer responses were recorded in the CRM database, and if the customer was uninterested, the reasons for the rejection would be recorded. If the customer was interested, the new product application was created and the transaction booked. The operations team was responsible for processing applications in a cost- and time-efficient manner.
If a service such as a loan or credit card was accepted, the SAU would immediately and rigorously start tracking for delinquencies; those results would be incorporated into the criteria, and the credit and SAU departments would review results. The SAU engaged with all of Dunia’s functional units and was empowered to make decisions and implement them. Cross-sell response results would be used to develop new models for predicting propensity to respond to similar future offers, and risk results were used to model and recalibrate existing behavior scores and other rule-based criteria. “It is not just about quantitative techniques, but also business sense,” Hurbas said. “One must know what to analyze depending on the problem at hand versus blindly repeating analyst reports at some frequency—then it becomes MIS and not analytics, which is really actionable MIS.”