Technology & Innovation

Spark of life: Maximising urban efficiency through location-based analytics

October 16, 2015

North America

October 16, 2015

North America

As one of South-east Asia's leading telecommunications firms, Singtel is no stranger to dealing with massive amounts of rapidly changing information. Recognising the untapped potential of this consumer data, the company created DataSpark, an analytics subsidiary, to provide insights to clients.

A new growth area for telecommunications firms

Facing increased competition from low-cost digital applications, major telecommunications firms like Singtel are seeking new sources of revenue. More than one hundred billion data points pass through the Singtel network every year. Executives recognised that all that traffic—which enables voice calls, data transmission and SMS messaging—provides an opportunity to build out an externally facing analytics business.

Singtel has been developing data-driven capabilities through an internal R&D unit since 2012. In 2014, the company created an independent spin-off, DataSpark, to offer insight services to external clients. It employs a proprietary geoanalytics platform to synthesise large quantities of mobile network data into discernible patterns of consumer movement. DataSpark extracts insights from this data to provide both public- and private-sector clients with a new window into the people movements that shape the services they demand. And Singtel uses leading-edge encryption algorithms for anonymising mobile phone location data, enabling them to extract key insights while safeguarding customer privacy.

Data for a purpose

This new organisation was charged with a dual purpose. In addition to brokering its unique brand of geoanalytics to clients in many different sectors, it was also tasked with helping the parent company to optimise operations in areas such as 4G upgrades and provision of a renewable stream of market intelligence.

"A lot of organisations try to find analytics or big data solutions by looking over the data they have and trying to implement something, anything, to see what happens," says Ying Shao Wei, Chief Operating Officer of DataSpark.

"Companies all too often make major non-recoverable expenditures in building state-of-the-art big data platforms without a firm idea as to how they will use them to generate value for the company—essentially creating a solution in search of a problem. We take the other approach—we decide where we want to end up and work backwards from there. If we don't have the data we need, we carry out experiments and ask ourselves how it could be stimulated or generated. It works very well and has played a major role in getting us where we are today."

Experiment, implement, succeed

DataSpark faced three distinct sets of challenges in implementing this new initiative. First, defining the value proposition of the data-driven business. Second, assembling a data-experienced team and building the capability to meet the needs of major clients in both the private- and public- sectors. And third, building support from the higher levels of the Singtel hierarchy. Geospatial analytics is in its infancy as a science, and DataSpark's executives and engineers needed to demonstrate that measurable returns would be realised from the solutions they proposed. Establishing credibility became top priority.

The answer lay in the core methodology: demonstration through experiment. "We built our analyses to show where the benefits would come from, and in that way we were able to change behaviour wihtin the organisation," says Dr Amy Shi-Nash, Chief Data Science Officer at DataSpark. "And, in time, this experiment-based approach became central to our developmental style. The data we use, the products we build and the benefits we anticipate are all linked together through technical experiment and market valuation, thereby helping to shape the decision process. We embark on that journey together with our clients."

Taking the pulse

Singtel is demonstrating how data can shape the development of a city through geoanalytics. The company is already levaraging its suite of tools to help municipal governments make smarter choices in their urban planning and transportation networks. In one notable success, Singtel conceived and implemented a real-time measurement system to assess the movements of Singapore's subway travellers anonymously to help the city's mass transit rail system better serve its 5.4m residents. Planning authorities use this information to optimise decisions about system mangement and future investments.

Singtel is also tapping into the value of geoanalytics for its private-sector clients. For example, it can determine the number of people who have passed by an actual or potential retail location by time of day, age, gender and other demographics. The company is beginning to experiment with indoor geolocation technologies that could help to map consumer movements within stores. "We are starting to offer a lot of different services, from optimising the locations of retail outlets and signage to increase visibility and attract target demographics, to matching opening hours and staff counts to actual traffic," says Dr Shi-Nash. "We've seen some early success in the retail space in Singapore and with our partner telco in Thailand and we're now developing our capabilities with another partner in Indonesia."

Singtel is also applying insights from DataSpark in its own operations to add intelligence to the mobile network, provide personalised customer care and place advertising. And new private, public and internal applications are being discovered every day.

From delivery to discovery

Helping to change the shape of Asia's major cities and consumer culture is only scratching the surface of this technology's potential. As the science of geoanalytics matures—and as the soures of accessible data deepen to include machine-data and the Internet of Things—Singtel plans to expand its portfolio of services. And naturally, the company also plans to further develop the sophistication and ease of use of its user interface—not only to help clients adjust to using these new analytics tools, but to broaden the possibilities of the science itself.

"The purpose of advanced analytics is changing from one of delivery to one of discovery," says Dr Shi-Nash. "It's very mimportant for users to be able to discover and ask the right questions, and to be able to find unexpected answers, in an intuitive way. Intuitive, user-driven discovery is a big driver for our insight offering."

Enjoy in-depth insights and expert analysis - subscribe to our Perspectives newsletter, delivered every week