A couple of main reasons are: 1. Accumulate a large volume of data where its value is based on sheer size. Their "revenue model" could be a red herring for a l

A couple of main reasons are: 1. Accumulate a large volume of data where its value is based on sheer size. Their "revenue model" could be a red herring for a later acquisition based on the monetary value of their datasets. 2. Analyze the data in-house and sell predictions, not the data itself which could skirt some privacy laws