Advertisment

Making Big Data Bankable

author-image
DQI Bureau
New Update

The financial services industry is more than just smitten by big data, there is no doubt. An eminent IT industry analyst says that in the 12 months up to October 2012, they received the maximum number of big data-related enquiries from banks. Research by another consulting firm specializing in financial services says that over 60% of banks believe that big data is important to their business. Little wonder then, financial institutions have a fine appreciation for and relatively clear understanding of the value inherent in their big data.

Advertisment

Protection against Fraud

A powerful early example of how this has helped them seize the advantage in today's competitive environment is tied to their ability to identify suspicious activity in real time to provide their customers better protection against fraud than traditional ‘post facto' solutions, which close the stable door only after the horse has bolted. With fraud detection and risk management topping their priority list, banks naturally gravitate to the possibility of using data from interactions on alternative channels, like mobile or social media, to spot deviant behavior or risk triggers, in real time, no less. And since we're not talking just credit card fraud-big data-driven fraud analytics can analyze and ‘risk score' structured and unstructured customer interactions as well as activity logs to prevent even account takeovers or insurance fraud. Last year, when I first demonstrated to banking leaders how account takeovers could be detected and prevented in real-time harnessing their big data pattern from historical transactions-the enthusiasm of this risk-intolerant financial community was quite understandable.

The Potential in Big Data

Advertisment

But, banks know better than to focus solely on the defensive aspects of big data, realizing that they would then be missing the forest for the trees. For big data has the potential to drive better, faster decisions across banking operations, not just fraud prevention, but deliver measurable value for their business. Going forward, specialist big data platforms could well provide all the answers banks seek. The innate capabilities of the platform-connectivity, scalability, compute power, visualization, institutionalization, and ubiquitous access-will enable financial institutions to straddle the data and analysis value chain end-to-end, from data discovery, to insight generation and decision execution. This means that the platform could easily allow them to gather big data in all its formats, varieties, and structures (or lack thereof), from each and every source-including hitherto untapped sources like conversations across the teller counter or from very large stores of unstructured and semi-structured data tucked away in their document management systems. In this way those walls in banks' current data infrastructure will finally crumble.

Even a year or two ago, whenever a bank needed a new analytical application, the business had to wait months while the technology team built it. On the other hand, these platforms can very easily be pre-configured with a repository of frequently used algorithms and reporting formats for business users to tap for generating insights. In fact, it is perfectly feasible to endow the platform with a tool enabling business users to create additional algorithms by simply dragging and dropping these on to a visual interface. What this means is that the bank's business users could be looking at time-to-insight within days, hours, or even real time. With that kind of capability, the bank's business users and IT enablers won't fail to realize the immense potential to be unearthed.

Anyone with an inside view will tell you how hard it is to get decision makers from different functions to arrive at one place, let alone at a consensus. Once again, the platform could resolve this quite easily by providing a ‘collaboration wall' on which executives from different functions and regions could exchange insights, interact with each other, and arrive at collective business decisions. It needn't stop at that. By equipping the platform with integrated workflow capability, it is possible to upgrade it into a best-action-recommending engine, which sees decisions right through to implementation.

Advertisment

An Example...

No doubt, skeptics will be tempted to dismiss this line of thought as far-fetched. Let's counter that view by casting the platform in a plausible scenario. In some developed countries, automobile customers prefer to change cars when their vehicle insurance is up for renewal.

Spotting an opportunity, insurance companies have entered into a quid pro quo arrangement with automobile manufacturers and their dealers-information on prospective buyers in return for a ‘fee' or a nice ‘little extra' that the insurance business can pass to their customers should the new car be purchased. While this might sound like a simple enough ruse that targets customer delight, accurately identifying ripe customers calls for processing massive amounts of data-structured and unstructured, both within and outside the business-big data to be precise. Imagine the time, effort, and money insurers could save if the platform were to make it easy for them to not only discover opportunities to delight customers, but also slot them into different categories or segments, and then kick off a series of steps to proactively engage with them.

That's but one example. Banks have applied that logic equally to common but key situations like, "Whom to target the new home loan at?" "Whose kids need a student loan?" and "What are customers saying about the bank?" After all, who better than banks to make big data bankable?!

Advertisment