When BNZ’s new customers reach their six month anniversary they become eligible for a “Health Check “ outbound call via the Customer Contact Centre focusing on the customer’s banking requirements and cross selling of new accounts.

However there was no systematic way in which eligible customers were selected for calling. BNZ needed to be able to identify which customers should be targeted as being more likely to open a new account.

The concept of a propensity model to rank leads was conceived to enable delivery of the 2011 business objective, to increase efficiency in how leads were routed and identify which customers were more likely to open an account.

BNZ had 18 months of contact and cross-sell data to model what were the drivers of positive responses. The model evolved over three to four iterations with validation and testing to ensure the strength of the attributes used. 

Essentially what BNZ did was turn historical data in to a predictor of future response, using a combination of profile and behaviour metrics to truly understand what drove its customers to respond.

By taking their thinking to the next level they unlocked the ‘why’ of customer response, not who responds. They were also committed to a constant evolution of learning and applying this to their campaign delivery.

This was not a one off solution but embedding a practice and a discipline.

BNZ took data and using sophisticated data modelling techniques turned it into insights and then into actionable results which generated clear and measurable improvements in financial results and increased return in marketing investment by an impressive 20 percent.