How Credit Card Companies Are Evolving with Big Data
As overseers of our digitized marketplaces, credit card companies have a bird’s eye view of what we buy. If you want to know what American consumers are interested in, there’s perhaps no better way that to examine their purchase histories, so it’s no wonder that credit card companies, such as American Express, Capital One, JP Chase Morgan, and Citibank are at the forefront of big data mining.
One of the credit card companies doing a lot of work in this area is American Express (NYSE: AXP). The 166-year-old company is a big adopter of big data tech, such as Hadoop and machine learning algorithms, to give it the data storage and computational heft needed for data mining at a massive scale.
The $34-billion company selected a Hadoop platform from MapR Technologies to store data and run big data processing for activities such as fraud detection, customer acquisition, and recommendation. With more than $1 trillion in annual transactions accounting for about one-quarter of all credit card transactions, American Express has lots of data to work with.
Detecting fraudulent transactions is arguably the biggest use case for big data at Amex, as it is for most financial services companies. Try as they do, it’s quite difficult for fraudsters to create transactions that totally mimic real transactions in every detail, and machine learning algorithms are quite good at picking out these anomalies.
But Amex has gone beyond fraud and is getting more creative with its big data tech, including recommending products and services to its customers.
For example, if a particular cardholder frequents a particular type of restaurant on a regular basis, Amex may recommend that the customer try similar restaurants in the area. It also has Amex Offers, which uses geographic information about a customer’s location to push offers to them in real time. The company, which recently launched a research laboratory in Silicon Valley, is also ramping up the use of big data tech in its merchant services division to sell anonymized data about cardholder transactions.
Hadoop plays a major role in Amex’s big data activities. “The Hadoop platform indeed provides the ability to efficiently process large-scale data at a price point we haven’t been able to justify with traditional technology,” Sastry Durvasula, Vice President and Global Technology Head of Information Management and Digital Capabilities within the Technology organization at American Express, says in a 2014 interview with ODBMS.org.
Big Citi Data
Citibank is even older than American Express. In fact, at 203 year’s old, it’s older than about 90 percent of the countries in the entire world. And judging by the New York City-based Citigroup (NYSE: C) subsidiary’s investments in data science and big data tech, it’s preparing itself for the next 200 years.
With more than 200 million accountholders in more than 140 countries, Citibank has a global reach that few corporations can rival, and with that reach comes an enormous amount of data to organizes. According to a 2013 SlideShare presentation by Juan Huerta, the VP of Advanced Analytics in the Global Decision Management business at Citibank, the company is taking a staged approach to big data.
That approach begins with establishing common data set and common tools and techniques for accessing data, according to Huerta’s presentation. At the sharper end of the stick, MapReduce-based machine learning algorithms play heavily in Citi’s Hadoop big data play, according to Huerta’s presentation, which also mentioned specific use cases in risk and fraud detection, contextual marketing, customer action, and clickstream digital.
Once the company’s data scientists create an analytic solution, the company aims to apply them globally. “We aim to offer our clients a level of service that is superior to other banks,” writes Don Callahan, head of operations and technology at Citigroup, in a 2013 piece on Financial Times. “We are focused on predictive analytics, learning not only people’s needs but what they are likely to want in the future–so we can be ready with the right products and services. We are also vigilant in protecting data and commit quite a bit of resources on information security.
Capital of Data
Founded in 1994, Capital One (NYSE: COF) is a relative newbie to the credit card business. But perhaps more than its competitors, the Tysons Corner, Virginia-based company is embracing technology as a way to differentiate itself in a marketplace that’s rapidly evolving.
According to a 2014 CapGemini Consulting case study on Capital One, the bank has been data-driven from day one. Instead of using a one-size-fits-all approach to pricing credit, the company gathered all available infuriation to make a customized offer to customers. “Essentially, what we were doing in the ‘90s was leveraging the power of data to custom-tailor products to our customers,” Capital One CIO Robert M. Alexander says in the case study.
Today, analytics are a cornerstone of the $25-billion company, which uses them not just for pricing and fraud detection, but also for predictive sales, driving customer retention, and reducing the cost of customer acquisition.
Machine learning algorithms play a critical role at Capital One. “Every time a Capital One card gets swiped, we capture that data and are running modeling on it,” Capital One Data Scientist Brendan Herger says in a video posted by H2O, which makes its machine learning software.
“The same thing [happens] every time we get a credit card application or a home loan application or someone calls into the call center to let us know that something is going well or poorly,” he adds. “All this gets captured and stored.”
The results of the analytics power decision-making at Capital One, and have made their way into new offerings, such as the Mobile Deals app that sends coupon offers to customers based on their spending habits. It has also enabled predictive capabilities in the call center, which CapGemini says can determine the topic of a customer’s call within 100 milliseconds, with 70 percent accuracy.
Analytics Sharing at JP Morgan Chase
Like most credit card companies, JP Morgan Chase & Co. (NYSE: JPM) has data science laboratories where technologists can play with new and emerging tech. But the 216-year-old banking giant has taken it one step further by analyzing accounts for clues about macro-economic conditions, including income and spending patterns, and sharing the results with the public.
Diana Farrell, President & CEO, JPMorgan Chase Institute, describes the impetus of the institute. “I can’t tell you how frightening it was to be in the middle of the debacle of the recession and not have a good understanding of what was happening in the household sector,” said Farrell, who was an economic adviser to President Obama during the end of the financial crisis in 2009 and 2010. ” We were just starving for real-time information.”
The NYC-based company conducted its research by tracking the spending and income patterns of 100,000 randomly selected individuals from a sample of 2.5 million accounts at the bank over a 27-month period December. The $96-billion bank discovered, among other things, that 40 percent of individuals saw their income vary by 30 percent or more from month to month, while 60 percent of individuals saw their spending vary that much.
JP Morgan Chase also discovered that Monday is the top spending day of the week and Sunday is the lowest. Americans spend three times as much on a Monday as they do on Sunday, they discovered. Not all big data discoveries are earth-shattering, but they invariably help fill in the gaps in our understanding of the world.