Big data puts analytics dream within reach

The problem of turning data into real value has been with us since ERP allowed businesses to record and store masses of transactional and operational data. While this promised the ability to revamp business, these systems simply turned paper-based logs into digital systems of record—albeit essential systems to enable efficient management of modern businesses.

But getting significant insight from this amount of data remains a challenge even with the emergence of data mining, business intelligence and analytics tools.

What's been missing is the ability to analyze this data at high speed from a wide range of data sources—and be able to associate data to a level that enables real intelligence and relevant business insight.

The emergence of technologies such as in-memory computing, high performance analytics and cheaper hardware have finally given businesses the tools to address this information challenge.

Next big thing?

Despite the hype of IT vendors (and the media), big data isn't as much of a hot button as many might think. During IDC’s recent Asian Financial Services Congress, Donald Macdonald, head of Group Customer Analytics & Decisioning at OCBC Bank in Singapore, revealed that searches for “analytics” has dwarfed searches for “big data” over the last five years. And that gap has not improved much even during the last year.

A 15-year industry veteran, Macdonald views the emergence of big data with a hint of skepticism. In his world and the talk is about analytics—how to derive the intelligence to improve customer value, improve operations and process, and improve the level of insight that can be gleaned from the masses of data that OCBC is able to access and leverage.

“There’s no doubt that big data in the examples of Facebook and Tencent offers huge opportunity to transform business,” said MacDonald, “but most businesses are a long way removed from those scenarios.”

Big data concepts

Those companies are pulling together massive data loads of structured and unstructured data—processing and analyzing the data in real time to deliver new levels of customer insight. For most businesses, that level of big data analysis is still far off, but what businesses are doing today is finally centralizing much of their data for analytics purposes, according to MacDonald.

The challenge is that the volume and variety of data is unsuitable for traditional databases and analysis techniques. The new tools at the forefront of big data hype now provide ways to address the volume, variety and speed of data which can stymie analysis.

Rather than discuss big data concepts with users and business leaders, MacDonald stresses the need to center discussions on desired outcomes around information, business insight, and questions on what data the business wants access to.

“It’s best not to get caught up in the technology,” he said. “Instead, focus on serving customers better or improving internal operations though better data or improved risk analysis—it’s not about being great at doing big data, but more about outcomes you achieve.”

Central intelligence

To achieve this, OCBC has had a strategic analytics vision for a number of years - there is now a centralized analytics team that sits within the bank’s customer experience group.

Over $100 million has been invested into analytics in the last eight years, and the analytics team is now a heavily used resource that serves all business units, except for risk-related operations which have a separate dedicated team.

“We are a center of excellence incorporating customer analytics, market research, customer experience design, as well as cultural change teams” said MacDonald.

He added that the challenge is to position this capability as close to the business as possible, but stay centralized so that the team can be proactive in applying analysis across the whole group. The objective is to discover insight and apply it across all product units—not just in isolation.

Interestingly, this team does not sit within IT as it does in many other companies. Macdonald sees both models but he observes that: “in general, when we see analytics within IT, the role becomes a basic order-taking role and becomes reactive rather than proactive. Putting analytics in IT also creates a disconnect – leading to analytical studies that are just not relevant to what the business is trying to achieve.”

Speed and context

According to Macdonald, the group is the cornerstone in driving cross-selling and up-selling activities, which now account for 30% of all credit card revenue, 25% of wealth management, and up to $85 million in shadow revenue.

The expectation for OCBC is to leverage emerging big data technologies to deliver even faster speed to insight and raise contextual insight and relevance of data findings. “Being able to create association between data sets - the ability to glean relevant insight across data sources and drive real-time decisions - is the key difference today,” Macdonald added.

One company focused on this ability to associate is Qlikview, an emerging player in the analytics space. “Businesses are seeking ways to get beyond a linear approach to BI and analytics,” said Terry Smagh, VP of sales of South East and North Asia, Qlikview. “Big data has helped drive this discussion and created new possibilities,” he added, “but the key difference from past experiences with BI and analytics is the ability around association.”

Being able to analyze across data sources and make relevant inferences, and to do this at speed, provides businesses with significant value and new revenue potential.

Candy Crush data crunch

Qlikview has customers like Crown Worldwide that leverage its technology to drastically reduce report generation (from weeks to minutes) which now enables real decision support. Retail and finance companies are tapping into new capabilities to deliver rapid business insight into customer trends and behavior, as well as critical supply chain information.

The makers of the gaming phenomenon Candy Crush,, leverage Qlikview to process 250 million rows of data a day for analysis into player usage trends.

Smagh likened the big data situation to the telecom “last mile challenge”: where people can see high speed fiber being laid in the street outside their house, but getting the last few feet into their own door is a problem. “Companies sit on masses of data, they store it, query and analyze it but are still struggling to unlock greater value from it all,” he said.

Today big data is happening—and whether people actually call it “big data” or not is irrelevant. Governments like Hong Kong use analytics to visualize public sentiment, and address complaints or issues that are discovered through e-mails and phone calls.

Further afield, examples like Citigroup show how big data is really pushing the boundaries. Tapping into IBM’s Watson artificial intelligence and analytics technology helps Citigroup to address areas like cross-selling as well as real-time credit risk and fraud detection. In Asia, the group has over 250 people working on data analytics. Last year, they opened an innovation lab in Singapore that combines analytics expertise with its key customers and connects them to another analytics hub in Bangalore.

Everything in business today is moving toward data being the driver for decisions and new business gains. Macdonald at OCBC said that no one in his company today can propose new plans, actions or initiatives without being armed with clear analysis and data to back up their words.

He noted that his world is full of many that say, “in God we trust, but please bring us the data.”