This is party why nearly all popular websites are based around collecting data about their users. They use this to make a profit and they also expand their services to include communications between their users which results in more data to analyze and collect. The gov't just needs to tap the business and rarely has to collect raw data on their own.
This isn't to say the raw data isn't being collected midstream, it's just that they sort of enable a distributed network to do the general analyzing before they focus on one target or group.--
Say no to JAMS!