Big data makes a big difference to banks

'Big data' is one of the latest buzzwords in the lexicon of IT professionals, especially in financial services. Banks and other financial institutions have always had to handle high volumes of data – concerning their customers and the products and services they use, and all the support functions. The amount of data available to them today, however, is truly mind-boggling.

Better and faster processing and storage is the reason for this explosion of information. Every expert has something to say on the scale of the growth, but a commonly quoted estimate is that the amount of data available to financial institutions is doubling every 18 months. That is forcing banks to invest in a new generation of platforms, tools, methodologies and skill sets.

The increasing level of information presents banks with great opportunities. It provides them with a deeper insight into customer behaviour and profitability. It facilitates the creation of new products and services. It helps executive management control the finances, measure and manage risk, and much more. But it presents problems too.

Old tools and technologies for storing, managing and analysing big data do not suffice. Creating new data governance and management policies is a must. Implementing these policies is costly, because it is a transformational process. The volume of data is in many cases unmanageable and banks have to decide what they want to keep and what they want to ditch. Even what they keep is not that useful unless it is structured. Most data is unstructured, meaning that it does not have a predefined data model, it does not fit well into relational tables and is usually text heavy, and text is harder to structure than numbers. But it is becoming easier to make sense of both structured and unstructured data.

Maximising opportunities

“Without question, big data provides financial institutions with a great many opportunities,” says Sharad Kumar, director of financial services at IT specialists, EMC Consulting. “It allows them to get closer to their customers by understanding their behaviour, which can be derived from deeper analysis of data from multiple sources and channels. One step ahead, it can help them predict certain events such as customer attrition or fraudulent behaviour. Big data analytics help them identify their more profitable products and monetisation of this data – selling the data and/or analytics on the data – presents them with new revenue opportunities. From an operational perspective, it helps them become more efficient and provides them with an ability to better manage risk," he says.

“Online retailers such as Amazon are among the best users of big data. They use it to find out an individual’s buying patterns and preferences, improve customer satisfaction, cross-sell and detect fraud. Banks, investment managers and insurers have seen what the likes of Amazon are doing and applying it in their markets,” says Mr Kumar.

Joe Dossantos, director of enterprise information management at EMC Consulting, says financial institutions have to act quickly to grasp the big data opportunities, even if the initial up-front investment is high. “If you do not act because you are worried about the costs, then someone else will step in,” he says. “You will save money in the short term, but in the long run the cost of inaction will be that you allow your competitors to move ahead.”

EMC Consulting advises financial firms, their IT experts and business executives on how best to manage and make commercial use of big data. IT departments want to know about the practicalities of gathering, cleansing, analysing, storing and retrieving the data. The business lines, on the other hand, are less interested in the technicalities and more interested in how they can use the data – especially the power of predictive analytics – to understand customers and their needs, to develop new products and services, and ultimately to monetise the data.

“We can help [financial firms] integrate the analytics into their upfront business processes,” says Mr Kumar. “It used to be the case that only the back office was interested in predictive analytics. Now they are used in the front office too, to help with direct customer interactions, dealing with mortgage applicants, underwriting insurance policies and giving investment advice. Analytics can drive decision-making at the front end of the business, as well as doing the number crunching in the background.”

EMC runs a big data advisory service (BDAS). “When new technology arrives, the non-IT people are often unsure how to use it,” says Mr Dossantos. “So the BDAS matches the technological developments to the business opportunities. Things such as Apache Hadoop – open-source software that allows big data to be stored and processed on a large number of computers – and Apache Hive – an open-source, big-data warehousing system than runs on Hadoop – are too technical for most business people. They only want to know how to use the technology to meet their commercial needs, which is what our BDAS shows them.”

Growth forecasts

Gartner, a Connecticut-based IT research firm, predicts that IT spending across all sectors of the world economy will increase to $3700bn in 2013, up 3.8% on 2012 – and it is spending on big data that is creating the most excitement. “By 2015, 4.4 million IT jobs globally will be created to support big data, generating 1.9 million IT jobs in the US,” says Peter Sondergaard, Gartner’s global head of research. “In addition, every big data-related role in the US will create employment for three people outside of IT, so over the next four years a total of 6 million jobs in the US will be generated by the information economy.”

But there is a problem. There are currently not enough skilled technology professionals in the US to cater to the big data demand. Mr Sondergaard says the public and private education systems cannot cope and, on current projections, only a third of those IT jobs will be filled. “Data experts will be a scarce, valuable commodity. IT leaders will need immediate focus on how their organisation develops and attracts the skills required,” he says.

As for financial services firms, Gartner published a report recently noting that banks’ big data initiatives are in their infancy, and for them to mature, banks must rewrite their data governance policies. “Many traditional data governance and data management practices are not capable of handling the deluge of big data,” according to the report. These practices must be standardised, transformed and integrated.

Commercial value

Alan Grogan, director of corporate analytics and value optimisation in the product, sales and marketing division of Royal Bank of Scotland's corporate and institutional banking, is recognised as a leading expert on big data and its commercial uses.

“Since the dawn of technology, data has always been used by banks to better understand the key drivers of organisational value and risk,” says Mr Grogan. “Banking leaders increasingly understand that big data analytics – turning data into actionable knowledge and insight – should not be restricted to the credit risk function, but that it presents endless opportunities across all areas of a bank. At RBS, we use analytics to free up staff time, which they can then use to spend more time with clients.”

Mr Grogan says the commercial aspects of big data need to be front of mind. “You cannot look at data across only the three vs: velocity, volume and variety. You have to give priority to the fourth v: value.”

The amount of data available to organisations has increased tremendously, but the infrastructure and human capital required to collect and manage it requires a significant investment. To control costs Mr Grogan recommends avoiding complexity, and reusing data across different functions and platforms to create synergies. Banks should also question whether predictive analytics – applying statistical methodology and computing power to data to predict what might happen in the future – is always necessary. Simple, cheaper human analysis of data may be adequate.

Ensuring that big data is properly managed and used is crucial. “The first building block is to have a central analytics function that is allowed to join up and generate value across what is really a spider’s web of data across organisational units,” says Mr Grogan.

“Big data that is mishandled statistically or wrongly interpreted is a very dangerous thing. So its operating model is critical. The leading analytically enabled businesses have been through their own learning curve and are now using analytics to drive a great deal of systems and process re-engineering capturing better data, refocusing staff time on a value basis and increasing operational excellence and effectiveness,” he says. 

Perhaps the biggest challenge of big data analytics is ensuring that the IT department is able to facilitate the analytics unit’s needs. “There is little point in having a team of analytics experts who are acting as important consultants to the business if you do not give them the right tools or if you restrict them with bureaucracy and a meaningless analytics infrastructure. That would be like hiring a Formula 1 driver and giving him a bicycle with a puncture and no seat,” says Mr Grogan.

Improving customer insight

The main beneficiaries of the big data trend are probably retailers, but banks are not far behind, according to Voranuch Dejakaisaya, head of information technology at Krungsri Bank, one of Thailand’s top five banks. “All of this data improves our customer insight, allowing us to enhance service as well as offer the right products at the right time,” she says. “It is also useful for credit risk analysis, fraud detection and anti-money laundering measures.

The bank is in the early stages of tapping into non-traditional data and accumulating it at a high rate. Ms Dejakaisaya estimates that the year-on-year growth rate will reach 800 to 1000% compared with about 30% for traditional data. The bank has only just started using Facebook and has yet to move into social media or streaming data such as images, voice and video.

“Once we tap into that terrain, we are talking about petabytes of data rather than terabytes,” she says. The most important aspect of big data analytics is to ensure that it is properly aligned with the bank’s product and service offerings, so the data is used to improve the customer experience and anticipate customer needs.

“IT departments must be willing to invest in talented personnel as well as the technologies if they want to get the most out of big data,” says Ms Dejakaisaya. “There should be regular ‘road shows’ to relevant business departments, telling them about any newly discovered insights that could be translated into commercial opportunities. In some cases business managers may make specific requests for analytics that support their objectives," she says.

“Big data is huge in volume, comes in fast and is widely varied. The challenge for organisations is to translate it into genuine insights that create business value. To do this they must align people, process and technologies. High-performance analytical tools are needed to create rapid results. A governance framework is essential,” says Ms Dejakaisaya.

Greater customer convenience

For IT executives at DBS Bank, Singapore’s largest bank by assets, the most important thing about big data is the positive impact it can have on the customer experience. “It allows us to know our customers well, hear them well and service them well,” says David Gledhill, managing director and head of group technology and operations at the bank.

He cites the example of how big data has been used to reduce the number of times that automatic teller machines (ATMs) run out of cash. DBS used to have a manual process for determining when they needed topping up, but that process sometimes got it wrong and machines dried up. Big data was the solution.

“We got all the transaction data on all of our ATMs, 24/7 over a number of years and gave it to a business analytics company who do high-end analysis work,” says Mr Gledhill. “They put their best optimisation and forecasting engineers on to it, working with us, to see how we could use technology to keep the ATMs filled. It involved looking at the dynamics of the flow of cash at thousands of access points to figure out how optimally to get notes to those machines. Even with all the PhD-qualified analytics experts deployed on the project, it took a long time to figure out how to do it. But we did figure it out and reduced ‘cash-outs’ by 80%.

“We also reduced the number of trips to fill up ATMs by 10% and optimised the amount of returned cash – how much comes back if the machines are not empty – by 60 to 70%. The vans filling the machines are on global positioning systems, so the system knows where they are and how soon the vans will get to a machine. Overall, it is an amazing use of data, involving really advanced analytics.”

It was the UK economist EF Schumacher who wrote the best-selling book Small is Beautiful in the 1970s, a critique of conventional Western economic thinking and it received much praise. Small may indeed be beautiful, but not when it comes to data. In the eyes of today’s financial services practitioners whose job it is to collect, manage and analyse data, and for whom information is their lifeblood, essential for career and company success, big is definitely better.