Blogs

An industry vertical analysis: Telecommunications and big data

Development Manager, Big Data. Information Management, IBM

We love our smartphones. And regardless of the service quality and high rates packages, we mostly continue to like our telecommunications carriers in one form or the other. IBM’s love for mobile is depicted by the fact that mobile is itself seen as data instead of just merely a device or collection of devices, generating an estimated 5.2 million gigabytes data per day.

Telecommunications is trailing behind

tower graphic from bg.pngHowever, to put it bluntly, the telecommunications sector has simply been an underachiever in the big data and analytics space historically, often failing to monetize the vast wealth of data spanning locations, interactions, personalized preferences and more. Some of this may be due to the fact that, after exponential growth in the 90s, telecommunications saw consolidation in mergers and acquisitions over the past decade, not to mention the frequent legal frays, which continue to the day of this writing.

While service providers have started to scratch the surface of analyzing retention risks and social influence using call record, operational and social media data, the industry has just not yet harnessed the wealth of data available to create that enhanced 360-degree view of the customer.

Changing an industry with big data

The sector is privileged to have ready access to customer-generated big data. By tapping directly into that data, telecommunications companies can easily analyze mobile network usage, application usage, service levels, customer micro-segmentation based on location and preferences.

Data federation plays a pivotal role in moving closer to that 360-degree view of the customer. The telecommunications sector has numerous data repositories built to pull switch data, social media data and so on. The challenge is aggregating the data. This is where IBM Big Data federation can help balance the equation by virtually defining a unified version of the data, while the data resides at it’s original source, thus minimizing data movement.

Also, small startups are in a hurry to utilize data to improve quality of service, but may simply lack the funding to acquire infrastructure and skills for such analysis. The good news is that governments across the globe are supporting incubation centers in liaison with high technology companies to help these startups. The big data cloud will play a major role in this scenario where infrastructure and skill barriers are reduced, for example: IBM Bluemix Analytics for Hadoop provisions dedicated instances to jump start big data processing.

Traditional databases are designed for a world in which data is sparse—big data changes our view of storage with estimates of only 5 percent of digital data being structured. IBM Big SQL MPP SQL engine for Hadoop powers IBM BigInsights, which will be a game changer in rolling out big data analytic deployments. Big SQL brings in the accelerated processing of semi-structured and unstructured data for BigInsights Hadoop storage.

To see this demonstrated, take a look at Wikibon’s case study comparison of Big Data MPP (massively parallel processing) vs. Warehouse analytics for financial ROI. The case study positioned big data analysis or an Internal Rate of Return of greater than 500 percent, with breakeven in less than four months! 

MPPvsDW_Figure1.JPG

Diving deep with big data

It will only get more exciting as telecommunications plunges deeper into the big data space, driving innovations at every level of connection and playing a crucial role in the years ahead as the backbone of the big data value chain network. It is also largely the case that this sector is virtually pinning the e-business boundaries defined by network penetration in various geographies.

Though growth and adoption in the telecommunications sector has historically been slow, there is clearly movement now to catch up and, eventually, lead. While we anxiously await the next smartphone model launch, I ask you, my readers, to vote on the upcoming industry vertical topic of my next post. It’s easy: simply select which verticals you would like to hear more about, and the winning vote will be the featured industry in my next post.

Create your free online surveys with SurveyMonkey , the world's leading questionnaire tool.