Zinia used a variety of unsupervised modelling techniques to find natural behavioural segments of customers. E.g. lifestyle conscious versus value conscious customers. These segments were then used for personalising marketing, customer journey online and product bundle offering. For example who are the right customers to sell contents insurance to when they are doing a mortgage application. Once these models have been built, they can be deployed using Zinia via a champion/challenger test and monitor over time.

Zinia was also used here to look at how likely any customer is to take-up the product at any application stage and which factors influence the take-up rate for each customer. As you get further into the application, more variables are available for the lender, so the models rerun and provide better accuracy/results which is used in real time for prioritisation.