Verizon FiOS, which provides bundled Internet, TV and digital voice delivered via a fiber optic network, wanted to know more about subscribers’ usage behaviors so that the company could profile and segment customers to create tiered billing and upsell/cross-sell opportunities and new service offerings.
Before the data lake: Because Verizon FiOS was unable to track network usage across multiple dimensions, it couldn’t determine which subscribers were using the most data resources and which services were most frequently used (e.g., streaming audio, video, Internet). The company needed a way to correlate huge volumes of network and subscriber data in near real time.
With the data lake: The data lake ingested approximately 3TB/day or 300K files/day and the system generated usage reports every 15 minutes. Verizon FiOS was able to correlate network probe data, such as source and destination IP addresses across multiple packets with subscriber data to capture individual data usage.
Results: Verizon FiOS was able to increase revenue by creating tiered data plans based on customer usage behavior.