Singapore

The Big Read: In the business of Big Data, Singapore has built a cutting edge

Big Data Analytics
How can Singapore overcome its intrinsic limitations on land, human capital and natural resources in its pursuit to meet the challenges of the future economy?
Published: 11:50 PM, February 24, 2017
Updated: 7:59 PM, February 27, 2017

SINGAPORE — If data is the new currency in the brave new digital world, there are few places outside of highly-wired Singapore that can boast of such riches.

While knowledge may be power, Big Data is big money: A recent report by research firm IDC predicted that the Big Data and business analytics market would grow globally from US$130 billion (S$182.4 billion) by the end of last year to US$203 billion by 2020.

Committee on the Future Economy (CFE) member Chan Chun Sing noted as much earlier this month, when he made the point that data is a valuable economic resource that allows Singapore to overcome its intrinsic limitations on land, human capital and natural resources. Coupled with digital connnectivity, the rise of Big Data means Singapore should be optimistic about achieving greater success, said the Minister in Prime Minister’s Office at the press conference to launch the CFE report.

Mr Chan added that while every country is keen to jump on the digital bandwagon, what sets Singapore apart is its ability to bring together different stakeholders and execute plans well, and its reputation of trust and reliability.

To this end, Singapore has already launched plans to transform into a Smart Nation, with the Government committing to making more data available to the public as it seeks to get the private sector to play a big role.

In Singapore, the number of mobile phone subscriptions outnumber the size of its total population by about 1.5:1, based on latest official statistics. As such, few organisations here — be it in the public or private sector — possess more data than the telecommunications firms, and they have wasted no time in getting bang for the buck.

From an individual’s profile, commuting patterns, to what he or she is interested in based on online searches, telcos supply a copious amount of anonymised customer data to companies as well as government agencies. In turn, the firms use the data to make big business decisions — such as where to set up new outlets or what kind of products and services to offer, and to who — while government agencies tap on the information to improve policies, programmes as well as services and infrastructure.

Safeguards such as personal data protection laws are in place to protect privacy. Even though identities are scrubbed out, the info that telcos can sell to other private or public organisations is extremely valuable given the insights that can be gleaned from co-relating the different data sets.

All three telcos here — Singtel. Starhub and M1 — provide data analytics services to businesses. In fact, Singtel set up a subsidiary, DataSpark, in 2014 with its internal resources to offer analytics and intelligence.

A wide range of organisations are hungry for the telcos’ data, including the government, attractions operators, as well as companies in the retail, transportation, travel and hospitality sectors.

DataSpark chief operating officer Ying Shao Wei said: “Data analytics allow businesses to find out where and when the crowds go including their home and work locations; see in-depth profiles of their customers, not just numbers of people; and discover patterns in customer location behaviour.” He added that clients use the “derived intelligence about the movement of people, vehicles and freight” to understand their customers’ journeys, improve retail networks and plan outdoor marketing campaigns.

The data has also contributed to better urban planning and public safety services, Mr Ying said.

For instance, public authorities can better marshall resources to ensure public safety, via insights from tracking how crowds build up and disperse, and by detecting anomalies in human traffic flow. DataSpark has also analysed commuters’ travel patterns for government agencies. Cellular data has been used to track, among other things, the number of people on the platforms of MRT stations, and the flow of commuters transferring between different lines at MRT interchanges. The data could then be used to calculate train frequency to minimise waiting times during peak hours, or determine the direction of escalators at MRT stations to facilitate the movements of commuters.

Underscoring the treasure trove of data that telcos sit on, StarHub chief commercial officer Kevin Lim said: “Telcos have data that is truly big data as we have millions of customers and hundreds of thousands of households (that subscribe to services). This data includes profiles, preferences and movement. When correlated, they offer tremendous amount of insights into behaviour and trends be it movement or preferences.”

He cited an example of how a customer that owns a big retail chain in Singapore used big data on people’s travel patterns to determine where future stores should be located. The retailer also used the data to predict which locations would bring in the highest revenues based on human traffic. This information, together with rental data, could be used to help the retailer decide on the final locations of its new stores.

A property developer also used the profiles of shoppers to determine what kind of tenants its mall in the east should have. For instance, if there are younger shoppers, it could potentially have more cafes. This is done by pulling together anonymised data on consumers’ online search or shopping habits.

At M1, government agencies, which are among its customers, have sought insights to improve the accuracy of their planning and “validate their strategies to enhance citizen services”, an M1 spokesman said.

A retail mall, for example, also used the telco’s analytics services to find out the time of the day when human traffic is high, and which entrances, exits or foot paths do most visitors use. Using such data, mall owners or operators can come up with differentiated rental rates, the tenant mix, and optimise maintenance and cleaning resources, the spokesman added.

Around the world, companies are ploughing in more and more investments into data analytics.

A reported published less than a fortnight ago by EY and Forbes Insights found that over the next two years, more than half of the 1,500 global executives surveyed are planning to invest at least US$10 million in data and advanced analytics. Several international companies dabbling in this area have sprung up over the past decade including DataVisor — which helps companies detect frauds before they occur — and Cogito, which claims to enhance the emotional intelligence of phone professionals through behavioral analytics.

The growing industry has opened up opportunities for many companies and start-ups here, apart from the telcos. One of them is homegrown e-commerce retailer and distributor Y Ventures Group.

(An employee at e commerce company Y Ventures Group poses for a photo with a computer screen showing a word cloud on Feb 23, 2017. Photo: Jason Quah/TODAY)

Using data analytics, Y Ventures forecasts trends for retail brands across 28 online marketplaces such as Amazon, Lazada and Qoo10. For instance, if a particular clock brand wants to find out how its competitors are doing, Y Ventures will identify the top 20 best-selling clocks on each online shopping portal. From this, it will find out which is the best performing competitor brand and track its sales volume, as well as predict the demand for its products.

Mr Matt Pollins, a partner at international law firm Olswang, advises companies on complying with the necessary regulatory requirements in handling data. “Every business now relies on data — whether it is banks who use data to monitor for cyber security incidents, or television companies who use data to establish what kind of programmes people most enjoy watching,” he noted.

The trend has been driven by an increasing amount of data, and better analysis and processing tools. “There is more data than ever before because there are more ways to collect data – millions of online interactions every second, more connected devices than ever before, and countless sensors on those connected devices. There are better ways to analyse and process that data because of cloud computing,” he said. In the past, servers and systems lacked the scale and sophistication to carry out complex data analysis, he added.

Mr Pollins said that ownership and privacy are the key considerations when it comes to legal and regulatory issues surrounding big data. “Some data is proprietary, which means it is protected from copying by others without consent of the data owner,” he said.

“We see organisations monetising their proprietary databases in this way. Meanwhile, some data is personal data — which means that it is protected by privacy laws such as the PDPA (Personal Data Protection Act) in Singapore. Increasingly, in big data, we see an overlap between proprietary and personal data.”

Under the PDPA, various rules govern the collection, use, disclosure and care of personal data. Organisations may use such data, but only if it has been anonymised and any personal identifiers of customers stripped away.

WHAT’S IN IT FOR THE PUBLIC

While private companies are reaping the rewards from big data, consumers arguably stand to benefit too — more directly in certain sectors, such as healthcare for instance.

Fullerton Healthcare sieves through millions of medical transactions at public hospitals to flag any spike in cases of a particular chronic condition such as diabetes in geographical areas, for example. Working with private companies, Fullerton Healthcare would then suggest an intervention strategy such as an education campaign to try and reduce the number of cases. It did this successfully with a company, where 70 per cent of claims were from 10 per cent of employees who suffered from chronic diseases. After it designed a chronic disease management programme for the company, the number of claims dropped by 60 per cent, and led to reduced sick days as well, said Fullerton Healthcare Group Chief Information Officer Ted Minkinow.

Similarly, healthcare technology startup RingMD has collected medical data to improve services for patients. For example, it is currently using big data to improve the way its chatbot interacts with patients. Its CEO and founder Justin Fulcher said: ““We’re using machine learning technology to make our health chatbot smarter, so that she can provide instant, personalised medical information.”

The company has also collected medical data such as heartrate and body temperature, from tracking wearables that patients put on. Using this data, it generates insights which are then provided to patients, to help them improve their lifestyles and health.

As for customer service software company Zendesk, it has a variety of services that enable companies to gain insights from what their customers say or do. For instance, it can track how their customers navigate the websites, or analyse chat conversations to look at what customers are dissatisfied with.

In Singapore, it has 800 corporate customers spanning across industries such as education, retail, telecommunications and real estate.

Zendesk Vice President of Engineering (Asia Pacific) Brett Adam said: “We find patterns in big data and use it to predict outcomes, such as if a conversation with a customer will end up with him being satisfied or not, or answer their questions automatically before agents can get to the customers. Our clients also analyse the data to track responsiveness, whether they are doing a good job responding to their customers, or how satisfied they are with the services they are getting.” Their clients may also use the insights to improve their products or services, he added.

In the public sector, the Government is a big user of data analytics, with data scientists at the Government Technology Agency (GovTech) using data to solve a variety of problems. So far, this ranged from resolving the mystery glitch that affected services on the Circle Line, to helping the Housing and Development Board (HDB) improve its services.

(GovTech Data Science Division consultant Daniel Lim poses for a photo on Feb 23, 2017. Photo: Jason Quah/TODAY)

Dr Daniel Lim, a GovTech data scientist, said: “We work closely with other government agencies to identify problem statements, and apply data analytics to derive actionable insights that could improve policy and operations. We do this in several domain areas, such as healthcare, transport, social policy and enforcement operations, to drive evidence-based decision making in the public sector.”

While GovTech works with various government agencies, it has deployed data scientists at three government organisations: The Ministry of Social and Family Development, The Ministry of Manpower, and SPRING Singapore.

For example, SPRING has started since two years ago using data analytics to track how its various schemes are doing, in the hope of improving them. “We track different kinds of data for management reporting and governance. This includes resource utilisation and the take-up rate of different schemes for local SMEs. The insights we gain using data analytics tools facilitate ongoing improvements and policy reviews such as our business grants. Ultimately, this benefits our people to do their job better and the businesses we assist,” said Mr Wong Ming Fai, SPRING’s chief information officer, who is attached to SPRING from GovTech.

Earlier this year, HDB started allowing new home owners to book an online appointment for their key collection. This was done after GovTech worked with HDB to analyse over 90,000 emails, and found out there was a cluster of feedback related to key collection appointments set by HDB. Dr Lim noted that government agencies receive millions of emails from citizens, “which provide a rich treasure trove of information about their concerns”.

He said: “Text analytics enables public officers to make sense of all these feedback and obtain insights into key issues that matter to citizens. By doing so, we can respond to these issues more proactively.”

Following its collaboration with HDB, GovTech has built a prototype text analytics platform, which is being trialed at several government agencies to enable public officers to use text analytics in their work, without needing to know how to code.

Even for the telcos, there is scope to do more as technology advances, with the end consumers at the heart of the efforts. These include using big data analytics to improve and offer more personalised services to consumers in the multimedia and entertainment space, said Singtel’s Mr Ying.

“The larger telcos are investing in advanced capabilities to build a 360-degree view of their consumers using wide-ranging data sources and compute and storage technologies. They could also use such capabilities to partner or help other enterprises and government agencies to address their analytical needs,” he added.

CORRECTION: In an earlier version of this article, Mr Ted Minkinow was wrongly attributed as Fullerton Healthcare CEO. We apologise for the error.