This example shows how to analyse data to provide more accurate data to input to a WITNESS model. It also illustrates a powerful standalone analysis identifying cross selling opportunities for a company across different product ranges.
A significant amount of the profit made by an insurance company is made through "add-on" products rather than the core product. A Company wishes to increase the amount of successful "cross-sales" by analysing groups of customers. If a particular group of customers has responded well to a particular add-on product, the remaining members of that group, and any new members, will be the first to be targeted during a marketing campaign or simply during customer service calls.
In this case study, the core products are buildings and contents policies and the add-on product is a discounted home alarm system.
The Company uses its central database system to aggregate the customer database and generate macro-level summary records describing its customers, which fall into 30,000 distinct groups.
This data in summary now shows the number of customers in each group and the percentage of each group that buys the add-on alarm system. Throughout this analysis it is important to remember that these are macro level records representing groups of customers, not individual customer records.
When using WITNESS to model business processes such as this, it is helpful to analyse the behavior of different sectors of the customer base. It is quite usual to find that the likelihood of different groups to take up cross-selling opportunities varies, and can affect process timings - for example, young people might take longer to reply if they are the successful target market, older people might need a higher percentage of a large-print form to be sent, thus elongating the process. It is important to establish major relationships in order to apply proper percentages and distributions to the rules governing the business process model.