Subscriber segmentation for customized marketing, direct-to-home tv provider


An Indian DTH TV Service Provider wanted to run targeted 1:1 marketing campaigns for their 11 million subscriber database. The challenge was the absence of key variables such as demographics, needs, lifestyle in order to develop an actionable segmentation model.


6 differentiated segments were developed based on a survey sample of 5000 subscribers. This segmentation from the ‘donor’ group was extrapolated to the ‘receivers’ (millions of subscriber base) using Machine Learning. These 6 segments were further divided into 4 groups each based on their value to the client: high/medium/low and inactive


Each of the 24 cells was reviewed by the client and their specific needs were identified based on the past viewership and transactional behaviour. Specific offers/products were mapped to each segment basis their likelihood to subscribe to them. Customized media channels were used to deliver these relevant value-added offers 1:1 to each subscriber.


The segmentation exercise helped deliver a four-fold increase in conversions of value-added packs, as a result of this targeted sales strategy