Font Size:
Modeling of Complex Network Data for Targeted Marketing
Last modified: 2017-04-11
Abstract
Developing strategies for targeted advertising of existing customers is a common goal in many business sectors, with usual practice focused on identifying shared acquisition patterns of products based on ownership data.
We observe customers' behavior for multiple agencies within the same company, monitoring choices of specific products along with co-subscription networks representing multiple purchases.
Our aim is to exploit co-subscription networks to efficiently inform targeted advertising of cross-sell strategies to currently mono-product customers.
We address this goal by developing a Bayesian joint model for mixed domain data which adaptively clusters agencies characterized by a similar customer base, exploiting a cluster-dependent mixture of latent eigenmodels to describe multi-purchase networks.
An application to data from the insurance market is presented.
We observe customers' behavior for multiple agencies within the same company, monitoring choices of specific products along with co-subscription networks representing multiple purchases.
Our aim is to exploit co-subscription networks to efficiently inform targeted advertising of cross-sell strategies to currently mono-product customers.
We address this goal by developing a Bayesian joint model for mixed domain data which adaptively clusters agencies characterized by a similar customer base, exploiting a cluster-dependent mixture of latent eigenmodels to describe multi-purchase networks.
An application to data from the insurance market is presented.