Listen to this article
Mobile network operators sit at the intersection of a set of emerging technologies that include big data analytics, location-based services and M2M (machine-to-machine) communications that are helping to redefine not only their own business operations, but also those of their customers.
Some are already using big data analytics to combat fraud or predict when customers are about to defect to a rival, while others are offering analytical services based on M2M data to help their business customers optimise their business models and improve productivity.
Like financial services firms as well as the media and entertainment industries, whose basic businesses now revolve around the bits and bytes of computer code, telecommunications companies have been among the first to adopt big data technologies, say IBM executives.
“Telcos generate enormous volumes of data,” says Fred Balboni, IBM’s global leader for business analytics and optimisation.
Much of data are stored in CDRs (call data records) that typically include the name of the subscriber, date, time and duration of the call and, depending on the type of call, additional data including switch data, cell tower IDs, device identification (serial) numbers, as well as International Mobile Subscriber Identity and International Mobile Equipment Identity codes.
By analysing this data, telecoms companies can make much smarter decisions about their customers, says Mr Balboni. “It’s not about the data; it’s about what you do with it,” he says.
For example, by combining CDR data with insights derived from sentiment analysis, mobile operators also know what type of people pass by a billboard at particular times of day and what products or services they might be most interested in.
Working with IBM, one major network operator is planning to sell this information to a digital billboard operator that should then be able to garner higher revenues by tailoring its billboard signage to the profiles of passers-by.
“Telcos can use customer intelligence to distinguish between customers, offer different interaction experiences to different customer groups and, in particular, identify priority customers,” says Ram Mohan Natarajan, senior vice-president for business transformation at Firstsource Solutions, an India-based outsourcing provider specialising in services for telecom operators.
In particular, he argues that telecom operators can use research to predict the future value of customers, and to identify those who are most at risk of leaving.
Huge numbers of customer retention decisions must be made, and made quickly, to retain the right people at the right price. The insights offered by customer intelligence analysis can help operators make the right choice,” says Mr Natarajan.
Bharti Airtel, one of India’s largest mobile network operators, was one of the first mobile operators to mine CDRs with the help of IBM, and use the customer insights it provided in this way.
Bharti Airtel generates a remarkable 6bn CDRs a day. “What is exciting is that the technology now exists to process this volume of transactions,” says Mr Balboni. By analysing these records IBM was able to help Bharti Airtel predict customer defections and create offers designed to retain them.
Other mobile operators are using big data analytics to discover who are their most valuable customers, even when, on the face of it, they pay the same flat-rate monthly fees and therefore appear to be of equal value to a mobile operator.
Some customers may actually turn out to be what Mr Balboni describes as “queen bees” – subscribers with extensive networks of friends who they contact frequently. “If a queen bee moves off the network, she might take 100 other users with her,” he explains.
“Network fraud is another huge problem for telecoms companies, often resulting in millions of dollars in lost profits while causing undue strain on the network,” says Splunk, a big data start-up with headquarters in San Francisco. Splunk’s technology tools are being used to help detect patterns and fraudulent activity as it occurs by correlating machine data across various sources.
MetroPCS, a leading North American telecoms operator, is using Splunk to index data from firewalls, intrusion detection systems and web servers to identify network abusers and take corrective action – plugging a key source of lost revenues. “Splunk is the one place we go to find our heaviest 'users’ and heaviest ‘abusers’. Within the first month we terminated enough rate-plan abusers to pay for Splunk,” the company says.
Mobile operators such as Vodafone Ireland also face challenges managing their increasingly complex networks. The company has deployed Tellabs’ insight analytics services to analyse its network data and use the information to optimise network performance and provide a higher quality of service. It will also enable Vodafone Ireland to perform capacity management and root-cause analysis quicker and more cost-effectively, the company says.
Similarly one of Splunk’s customers in Asia is using machine data gleaned from across the network and from devices hooked up to it, including handsets and set-top boxes, to improve the efficiency and productivity of their network. The company can now detect network performance issues, the most downloaded content and the most popular requests, all in real time. These insights have helped them improve content quality while ensuring optimal network performance.
As a recent independent research paper published by Comptel, a specialist telecom software provider, noted, “communications service providers that have the ability to handle large data volumes in real time, undertake predictive modelling and automate decision-making and action-taking will realise better customer satisfaction, increased churn reduction, enhanced operational efficiency and increased revenue, among other performance improvements.”
When it comes to big data analytics, mobile operators have only just begun to scratch the surface of what may be possible in the future.
By sifting through the huge volume of data they collect in real time, and marrying that with external data sources such as demographic information, location related information and even weather, they may in future be able to offer a wide range of new services to subscribers and in doing so, head off the threat that they simply become what some have described as “a dumb – albeit mobile – pipe”.