Big Data: Telcos Can Capitalize on Home Turf

Telcos looking to cash in on the vast volumes of data at their fingertips need to focus on the local advantages they hold over their data mining-savvy Web services rivals.

That was a key message from Ubuntu founder Mark Shuttleworth as he addressed delegates at the recent Mobile Asia Expo in Shanghai.

He believes that compliance with local data laws and the ability to store applications and data and perform analytics functions locally too could provide communications service providers (CSPs) with a unique Big Data opportunity.

Shuttleworth believes the future for CSPs lies in the development of analytics-driven services. "The services you wrap around [the analytics] are unique to your industry and unique in many cases to your geography," so it makes sense to focus investment on supporting such advantages.

He urged operators to think carefully about where they site their application hosting and analytics capabilities, suggesting that it makes sense to host such IT systems as close as possible to end users. "In 10 years time, people will look back and say: 'Where you chose to put data is very material.'"

This opportunity, and the ability to win compliance from local regulators, are "the only advantages [telcos] will have in the new emerging race between incumbent operators and new kinds of provider like Google," he added.

Currently, churn forecasting and reduction are popular early Big Data projects for operators.

For example, SK Planet, the marketing subsidiary of South Korea's SK Telecom, is putting Big Data analytics to work to help its parent cut churn and to generate new revenue.

CEO Jinwu So told Mobile Asia Expo attendees that SK Planet's data mining had enabled SK Telecom to achieve a fourfold improvement in churn forecasting. According to So, users planning to churn tend to search for very specific phrases, such as "data plan" or "operator benefits," at least three to seven days before they take any action.

"They also tend to go to specific sites that provide data about carrier offerings and information on devices," added So.

SK Planet has also built a massive loyalty program, called "OK Cashbag," with 36 million members. Last year it generated US$50 billion in transactions and accounted for 10 percent of all Korean retail purchases. By combining its own data with those of its thousands of retail partners, it is able to provide much more targeted coupon offers and direct mail messages.

It also runs 11th Street, South Korea's biggest mobile commerce and second biggest e-commerce site. That site offers a daily deal service that has doubled its purchase conversion rate by applying an analytics-based strategy.

So said the key was to "build an eco-system not only for our own service but to work with partners, including retailers." This meant mining partners' customer data as well as SKT's, he added.

Not surprisingly, both Shuttleworth and So agreed that privacy is one of the biggest challenges. "We're still building a new social consensus where we can balance between privacy and the benefits of Big Data," Shuttleworth noted.

Speaking to Light Reading following his presentation, So said SK Planet had the permission of its users to use their data from Web searches and retail purchases. But he says privacy is a growing problem. "It's getting more and more difficult as the rules keep changing," he said.

For more on Big Data in the telecom sector, see:

– Robert Clark, contributing editor, special to Light Reading