Data Management: “Big data” at the bank

Back in the day, you got to know your customers by chatting with them across the teller line or on the platform. You got a sense of how their retirement plans were shaping up, what colleges the kids were planning to go to, or what kind of car they might want to buy. The good customer service representatives could sense how the bank could meet their needs.

Now, who comes to the bank any more? Most customers tap the ATMs, fire up their computers, pinch their smartphones, or maybe dial the voice response unit. At the same time, the bank likely is squeezed by pressures to drive revenue, cut costs, comply with mounting regulations, and fight fierce competition.

The era of “big data”

Many banks are coming to realize that the very technology that has physically distanced them from their customers holds the potential to allow the banks to know their customers—and potential customers—much more intimately, systematically, and economically than ever before. It’s the era of “big data.”

“Big data essentially is this proliferation, exponentially, of data about consumers from the banking perspective and every aspect of their behaviors and attributes. We see it in the form of new channels, with consumer behavior being captured in multiple ways, in payment systems, online banking, card usage,” says Curt Johnson, director/product management at Open Solutions.

The concept is simple, but execution is not. Part of it is technical and involves advanced technology. But another part—perhaps a larger part—involves the human element. Technology can’t be left to the CIO anymore. But the CIO needs to be fully plugged into the bank’s operational strategy and objectives.

Big data means really big. By one measure, 90% of all the data that is stored in all the databases in the world today was generated in the last two years, says Yi Deng, dean, College of Computing and Informatics at the University of North Carolina-Charlotte. “We’re talking about doubling the data volume almost every two years for the foreseeable future. That is a sea change in which people are in the process of absorbing and trying to figure out what to do with it,” Deng says.

Use big analytics, in real time

Data is also big in usefulness. “There has to be a business case, and there is,” says Peter Graves, CIO, Independent Bank, Ionia, Mich. “This would be to understand [customer] buying habits. In banking, it might be what products they have, how often they change banks, how often they refinance their mortgage. It’s taking every aspect digitally of that relationship. It may not just be what they have with you; it may be their entire financial picture. Then you can try to make sense of that and match products and services in an efficient way that meets their needs.”

There is a need for analytics—the tools to analyze data, sort out important nuggets, and recognize and ignore useless noise. That, in turn, leads to actually using the useful information in a timely way.

“The crucial thing for most banks is not so much the sophistication of their model or the quality of their data, although obviously that’s foundational. It’s their ability to operationalize around that, to do something different,” says Nigel Smith, managing director/banking distribution and marketing services in North America for Accenture.

Smith continues: “The big trend in analytics is getting it out of the back room, away from the [tech specialists], and making it operational in the front lines. It’s asking, what are we going to do differently in the way we design products, market products, and deliver those products and services to customers based on the insight we’ve gained?”

He suggests three examples. One: Test pricing sensitivity, comparing real-world experiences with the bank and competitors, to find out how to design different product and service bundles and associated prices that likely will be favorable to customers. Another: Optimize distribution, figuring out how to best use all the different channels for sales and service. A third would be to really understand who in the bank is best at sales, and then determine, in detail, what makes them so good and if those traits could be transferred to the other sales staff.

Jeanne Johnson, global leader/information intelligence at KPMG, says data collection and analysis can drive two separate but parallel processes in a bank’s operations: the information needed to complete a transaction, and the understanding needed to ensure the parameters of such a transaction are recalibrated instantly for maximum effect.

She provides another real-world example of big data/analytics:

“As a transaction, what information does an underwriter need to consider getting a loan approved? A lot of that is automated.… You need to know the information that’s captured in that first screen. Before that discussion ever happens, however, there are some credit guidelines that, as a result of reviewing historic information, may be combined with some predictive information about what’s trending.”

Johnson goes on to say that this results in some pre-established policy limits. “Those policy limits need to be constantly rechecked, revalidated, recalibrated based on all the lending activity that’s going on. [So] you can use an information data collection strategy to better inform on a real-time basis whether the policy needs to be changed.”

Take a strategic approach

Vendors and large-bank research departments are working on systems to handle and analyze big data, but pitfalls loom large. “The challenge is, number one, getting grips on all the data that you have at all the sources. The second challenge is to integrate it in a way that you can make sense out of it. Finally, then you have to be able to store that and do it in a way that’s user-friendly,” says Independent Bank’s Graves.

Making sense out of and using the data is where the human element plays the crucial role. “What a bank needs in place is a strategic roadmap for data: how to collect it, how to treat it, how to use it … so there is a structured approach to gathering the data. Otherwise you’re going to try to consume too much, look at too much, and risk not making sense of anything,” says Johnson of Open Solutions. “From a larger perspective … you need a person within the organization in charge of executing on that strategy.”

In many cases, it is the CIO. Success, says Accenture’s Smith, “depends on the ability of the CIO to partner across the executive team, with the chief marketing officer, with whoever owns distribution, and to find more pragmatic, tactical solutions with that insight and product analytics, and change how the organization works.”

“What we’ve seen in the past is a lot of silos within the organization,” explains Lizette Nigro, vice-president/core product leader, Open Solutions. “Commercial lenders, consumer lenders, branch retail folks. They would always go to the IT area and say, ‘I need this’ and IT would help find the solution for that. The information group of a bank really needs to be able to take all of those requests and tie them together and understand the larger need of the financial institution.”

Yi Deng, of UNC-Charlotte, says the business case of big data really doesn’t focus strictly on IT investment. “Managers [of any kind] have a stronger need now to understand the right questions to ask of the technology people,” he stresses. “It requires organizational change. It requires a culture change. It requires a broader range of innovation. And it requires a whole new group of talents in large numbers, which we don’t have today.”

Start with what you have

All of this sounds daunting, but the first steps don’t have to be. In fact, banks of all sizes have analyzed considerable amounts of customer data for years: It’s called cross-selling.

In a recent report, Curry Pelot, Fiserv’s CIO, describes a way for banks to drive sales using information they already have.

“A branch won’t be highly successful at trying to sell mobile banking to retirees living off fixed incomes,” writes Pelot, by way of example. “That’s only logical. But other realities aren’t as obvious,” he says. “How many of the branch’s customers are candidates for a mobile-banking product? Which ones are most likely to open an account? Based on that information, what’s a realistic sales target for each individual branch? What kind of dollar return should you expect from your cross-sales efforts?”

His five-step plan for success involves knowing current customers; measuring what they’ve already bought; determining the best cross-sell targets; setting realistic goals for each branch; and allocating the appropriate resources.