Many CSPs Have No Big-Data Analytics Strategy

By Craig Galbraith Comments
Print

More than one-quarter (27 percent) of communications service providers (CSPs) have no strategy in place for "big data analytics," a surprising revelation as we enter the second month of 2013.

New research from Analysys Mason shows that the vast amount of information CSPs hold about their subscribers has been largely untapped. The volume of data on telecom networks has increased by a thousand in the last 20 years, and more data has been created in the last two years than in the preceding 50.

“The data that CSPs are creating [have] four key attributes," said Patrick Kelly, lead author of the report and research director for Analysys Mason’s telecoms software research division. “The data has volume (there is lots of it), variety (from call logs to M2M sensor data, it is extremely varied), velocity (it can be gathered in real time) and value (if structured and analysed correctly, it can be extremely valuable and profitable)."

The report says CSPs better understand the outcomes for specific areas of their businesses before investing in big data and analytics. It's estimated that CSPs can increase their net profit margins by 12 percent with the right cross-marketing and sales promotions, and customer retention can be increased by 0.2 percent with effective loyalty programs.

Only a fraction of the data that traverses telecoms networks needs to be captured for analysis, Kelly said. Three types of data are important for CSPs: customer data (usage, location, device, etc.), market intelligence (dimension, demographics, segmentation, etc.) and real-time network data (service quality, call-center efficiency, revenue optimization, etc.).

The term "big data analytics" is associated with capturing and analyzing consumer behavior using the Web, and can have huge benefits in terms of both revenue generation and customer loyalty, Analysys Mason said.

Follow senior online managing editor @Craig_Galbraith on Twitter.

Comments
comments powered by Disqus