Genotyping of Digital Music Subscribers
This paper presents a methodology to categorize subscribers of digital music service (DMS) by taking as input variables their historic download pattern and streaming library. Drawing inspiration from biology, we develop a metric called "genotype" by defining a series of indicators called attractors and detractors that form a category space or "species" for every user. These species are based on four main styles of music: latin, urban, rock and pop; the indicators assign weights to the genres based on the sociological subjective perspective of music fans from one category in relation to other music styles, i.e., how they view other types of music they don't feel affinity with. The result is a segmentation of users that finds application in the making of offers and promotions, which can in turn be coupled with association rules and market basket analysis to improve direct marketing campaigns (CTR) and maximize revenue.