Big Challenges for Big Data
To that end, most communications service providers (CSPs) are starting small with an initial implementation of analytics tools to manage data volume. The tools look for simple patterns and obvious trends that humans would see if they had the wherewithal to process all the necessary data points. In short, these tools know what the answer looks like, and that’s a good place to start.
For example, if a large number of new customers in a specific area sign up with a CSP at a retail outlet, then perhaps the CSP needs more outlet stores or additional staff. And if the ratio of customers signing up for a CSP’s new service at a sporting event is better than the ratio at a mall kiosk, perhaps the CSP needs fewer kiosks but more staffing for large events.
Starting with these types of analyses across large data sets builds a user’s confidence in regard to the tools that are being employed and helps the user understand how to formulate questions to ensure that the analysis of Big Data delivers insight.
Slow start
A study conducted by IDC in 2012 claimed that only one half of 1 percent of data across the globe was being analyzed. But beyond the obvious challenges of capturing and correlating all that data, a sticking point for many service providers is their ability, or lack thereof, to take action in response to data analysis. Many CSPs are applying analytics to Big Data for the purpose of understanding network, not customer, behavior. Why? Because processes and systems are already in place to address network problems and improve network performance. The tools needed to react to many customer issues, on the other hand, just don’t exist.
If an analyst detects fraud, a connection can be shut down and future access can even be blocked. Conversely, if an analyst detects that a popular application isn’t performing well on a specific smartphone, there is no easy way to quickly identify affected customers and notify them, then determine if a fix is available and automatically upgrade the device in question over the air; the processes aren’t in place, and CSPs’ operational and business support systems (OSS/BSS) can’t handle this sort of problem. The tools may exist to tame Big Data, but service providers are hesitant to make the investment if they can’t act upon the findings.
Talent and technology are expensive, so it’s important to look closely at the business case for Big Data. If the back-end processes and systems that allow a company to proactively respond to the information being generated aren’t up to the task, costs soar. Access to unstructured data is difficult to come by and often requires expensive integration. Additionally, IT and other departments within a company aren’t always on the same page, while other departments (e.g., marketing) aren’t typically known to be process advocates.
