Advanced Analytics – Retail & Ecommerce

Business Challenge

In order to predict customer lifetime value, the client needed:

• To know how much revenue the customer is likely to give in the next couple of years based on early engagement with their games.
• To rank order the customers based on the predicted class they belong to so as to understand their contributions to the profits rather than treating all customers alike.
• To empower its decision makers by giving them the ability to extract and analyze customer data for informed decision making.

Features

At Intuceo’s core are five patent-pending analytic and predictive engines:

• Data SharpTM pre-processing wizard – the total revenues generated by customers was binned into 4 classes
• Quick InsightsTM preliminary insight validator – created initial variables – like does the family have a preferential game or do they play several games, what is the age differential between eldest and youngest kid in the family
• Hidden InsightsTM deep insight extractor – extracted about 82 attributes including the class attribute and these were used to build the models
• Bold VistasTM predictive modeler – predictive models were delivered to the business users via comprehensive self-explanatory graphs and charts on a web dashboard.

Benefits

• Accurate decision making with customer micro-segmentation – People who play a single game in the first week are likely to give 1/3rd as much revenue as people who play multiple games.
• Targeted marketing – Families with more kids means more revenue – giving $11 in revenue per year per kid.
• Improved customer retention – Customers who played multiple games are sticky AND give more revenue

Solution:

The client implemented iCube’s predictive analytics solution to predict the lifetime value of a customer based on his or her early engagement. Data set consisted of 6,093,598 customer transactions of usage data collected over 5 years for 10,000 unique customers.