In Plain Sight: The Extraordinary Potential of Big Data
The era of big data has arrived, and few industries are better positioned to benefit from it than banking and financial services.
Thanks to the proliferation of smartphones and the growing use of online social networks, IBM estimates that we create 2.5 quintillion bytes of data every day. In an average minute, Yelp users post 26,380 reviews, Twitter users send 277,000 tweets, Facebook users share 2.5 million pieces of content and Google receives over four million search queries.
Just as importantly, data centers have slashed the cost of storing information, computers have become more powerful than ever and recently developed statistical models now allow decision makers to simultaneously analyze hundreds of variables as opposed to dozens.
But while fintech upstarts like Simple, Square and Betterment are at the forefront of harnessing data to tailor the customer experience in their respective niches, no companies know their customers better than traditional financial service providers. The latter know where their customers shop, when they have babies and their favorite places to go on vacation, to mention only a few of the insights that can be gleaned from proprietary transactional data.
When it comes to big data, in turn, banks have a potent competitive advantage given their ability to couple vast internal data repositories with external information from social networks, Internet usage and the geolocation of smartphone users. In the opinion of Simon Yoo, the founder and managing partner of Green Visor Capital, a venture capital firm focused on the fintech industry, the first company to successfully merge the two could realize “billions of dollars in untapped revenue.”
Few financial companies have been as proactive as U.S. Bancorp at embracing this opportunity. Using Adobe Systems Inc.’s cloud computing services, the nation’s fifth-largest commercial bank “integrates data from offline as well as online channels, resulting in a truly global understanding of its customers and how they interact with the bank at multiple touch points,” says an Adobe case study.
By feeding cross-channel data into its customer relationship management platform, U.S. Bancorp is able to supply its call centers with more targeted leads than ever before. The net result, according to Adobe, is that the Minneapolis-based regional lender has doubled the conversion rate from its inbound and outbound call centers thanks to more personalized, targeted experiences compared to traditional lead management programs.
Along similar lines, a leading European bank studied by Capgemini Consulting employed an analogous strategy to increase its conversion rates by “as much as seven times.” It did so by shifting from a lead generation model that relied solely on internal customer data, to one that merged internal and external data and then applied advanced analytics techniques, notes Capgemini’s report “Big Data Alchemy: How Can Big Banks Maximize the Value of Their Customer Data?”
Another European bank discussed in the report generated even more impressive results with a statistical model that gauges whether specific customers will invest in savings products. The pilot branches where the model was tested saw a tenfold increase in sales and a 200 percent boost to their conversion rate relative to a control group. It’s this type of progress that led Zhiwei Jiang, Global Head of Insights and Data at Capgemini, to predict that a “killer app” will emerge within the next 18 months that will change the game for cross-selling financial products.
The promise of big data resides not just in the ability of financial companies to sell additional products, but also in the ability to encourage customers to use existing products and services more. This is particularly true in the context of credit cards.
“In a mature market, such as the U.S., Europe or Canada, where credit is a mature industry, it is naïve for a bank to believe that the way it is going to grow revenue is simply by issuing more credit cards,” notes a 2014 white paper by NGDATA, a self-described big data analytics firm. “The issue for a bank is not to increase the amount of credit cards, but to ask: How do we get the user to use our card?”
The answer to this question is card-linked marketing, an emerging genre of data analytics that empowers banks to provide personalized offers, savings and coupons based on cardholders’ current locations and transactional histories.
The venture capital-backed startup edo Interactive does so by partnering with banks and retailers to provide card users with weekly deals and incentives informed by past spending patterns. Its technology “uses geographical data to identify offers and deals from nearby merchants that become active as soon as the customer swipes their debit or credit card at said merchant,” explains software firm SAP’s head of banking, Falk Rieker.
Founded in 2007, edo has already enrolled over 200 banks in its network, including three of the nation’s top six financial institutions, and boasts a total reach of 200 million cards.
Poland’s mBank offers a similar service through its mDeals mobile app, which couples the main functions of its online banking platform with the company’s rewards program. “What makes this program so innovative is its ability to present customers with only the most relevant offers based on their location and then to automatically redeem discounts at the time of payment,” notes Piercarlo Gera, the global managing director of banking strategy at Accenture.
A third, though still unproven, opportunity that big data seems to offer involves the use of alternative data sources to assess credit risk.
The Consumer Financial Protection Bureau estimates that as many as 45 million Americans, or roughly 20 percent of the country’s adult population, don’t have a credit score and thereby can’t access mainstream sources of credit. The theory, in turn, is that the use of additional data sources could expand the accessibility of reasonably priced credit to a broader population.
One answer is so-called mainstream alternative data, such as utility payments and monthly rent. This is the approach taken by the VantageScore, which purports to combine “better-performing analytics with more granular data from the three national credit reporting companies to generate more predictive and consistent credit scores for more people than ever.”
Another is to incorporate so-called fringe alternative data derived from people’s shopping habits, social media activity and government records, among other things. Multiple fintech companies including ZestFinance, LendUp and Lenddo already apply variations of this approach. ZestFinance Vice President for Communications and Public Affairs Jenny Galitz McTighe says the company has found a close correlation between default rates and the amount of time prospective borrowers spend on a lender’s website prior to and during the loan application process.
“By using hundreds of data points, our approach to underwriting expands the availability of credit to people who otherwise wouldn’t be able to borrow because they don’t have credit histories,” says McTighe, pointing specifically to millennials and recent immigrants to the United States.
While this remains a speculative application of external data by, in certain cases, inexperienced and overconfident risk managers, there is still a growing chorus of support that such uses, once refined, could someday make their way into the traditional underwriting process.
This list of big data’s potential to improve the customer experience and boost sales at financial service providers is by no means exhaustive. “It’s ultimately about demonstrating the art of the possible,” said Wells Fargo’s chief data officer, A. Charles Thomas, noting that big data could one day help the San Francisco-based bank reduce employee turnover, measure the effectiveness of internal working groups and identify more efficient uses of office space.
It’s for these reasons that big data seems here to stay. Whether it will usher in a change akin to the extinction of dinosaurs, as Green Visor’s Yoo suggests, remains to be seen. But even if it doesn’t, there is little doubt that the possibilities offered by the burgeoning field are vast.