Banks and other financial services companies sit atop vast and potentially valuable stores of data, but most struggle with harnessing that data such that it can tell them something about their customer base, let alone sell them something.

But that hasn’t stopped a few companies from experimenting with it anyway.

The credit reporting bureau TransUnion announced recently that it will launch a new platform in April that gives lending institutions access to more than 350 billion rows of data representing seven years of trade line history on about 200 million American consumers.

Paul Siegfried, senior vice president and credit card business leader at TransUnion, said the company developed the platform, named Prama, in response to challenges expressed by its clients, mainly regional-sized banks. Large data sets can be difficult and time-consuming to manage and simply having the data isn’t enough to be able to draw insights from it, he said.

Siegfried envisions banks and credit unions using the data they glean from Prama to make decisions about entering new business lines or markets or handling customer accounts if they start to see delinquencies in a particular area.

“Being able to manage that data, being able to get the depth of that data, gives us more perspective,” he said. “That’s the power of big data. That’s why it’s important to banks.”

Closer to home, Eastern Bank’s tech incubator Eastern Labs recently launched its “Express Business Loan,” which asks small business customers just four questions and returns a loan decision in five minutes or less.

That loan product, the first out of Eastern Labs, relies on myriad sources of data, both publicly available and gathered from the bank’s own customer set, as well as the data science models Labs has built since its inception just two years ago. It’s a new and faster way of executing against the bank’s currently existing credit policy, and Eastern is hoping this product will help it better compete with nonbank online lenders that sometimes charge exorbitant interest rates.

“What I think is so fascinating is they’re both trying to leverage their competitive advantage with their data,” commented Traci Hess, a professor of information systems in the Isenberg School of Management at UMass Amherst.

“Of course, TransUnion is commoditizing their vast stores of data, but they don’t have the ability to experiment with their products as readily,” she said. “On the other hand, [Eastern Labs is] trying to leverage the competitive advantage that the data from Eastern Bank provides them. They get to experiment with their products more readily. They can use the Eastern Bank data but they can roll out products and have a ready test bed through Eastern Bank.”

“It’s fascinating to think about the different angles these companies are taking. They’re commoditizing different data sources and they each have an advantage,” she said.

Of Wild Stallions And
Sentiment Testing 

Eastern Labs and TransUnion aren’t playing with what we would exactly call “big data” in these cases. To be clear, “big data” generally refers to even larger data sets than what these companies are dabbling with, but both represent interesting new ways that financial services companies are capitalizing upon the vast stores of data available to them.

David O’Connell, a senior analyst at Aite Group, believes that if bankers do figure out how to combine elements of big data with sentiment testing, they’ll be able to glean much more timely insights about what’s going on with their borrowers and better respond to either risks or opportunities.

Bankers get financial statements and the like from their borrowers, but that information alone is not necessarily enough to assure a borrower’s continued strong performance, he said.

“I’m quite convinced that if you took every bankruptcy that occurred and looked at everything that was said about that company out on the web, you would find a very strong set of clues leading up to that,” O’Connell said.

He gives the example of a retailer that sells sneakers. If there’s a lot of negative Internet chatter about a particular brand of shoe, that could mean the retailer missed the mark and is selling a shoe that’s no longer popular.

However, if that chatter concerns delivery times and availability of the product, that might mean the borrower is selling out of stock fast. Armed with that information, the banker might call up the borrower in question and offer to increase their line of credit, for instance.

“This not only gives banks an opportunity to do borrower sentiment testing, to predict and do something about risk, but also to do something about favorable performance and be that responsive banker,” O’Connell said.

The financial services world still has a long way to go before it gets to that point. Right now, banks struggle with more traditional analytics, never mind big data, and most small to mid-sized financial institutions can’t dedicate a full-time data scientist to the job of organizing and sorting through said data to draw conclusions from it.

In the meantime, O’Connell recommends that bankers ask their potential vendors to not only automate certain tasks for them, but also create robust back end data sets that have lots of prebuilt reporting and the ability to integrate with other capabilities.

Or as Siegfried put it, “Big data is like that wild stallion running across the plain. You’ve got to catch him and bridle him and figure out how to control him so he’ll take you from point A to point B.”