‘Big Data’ Offers Potential for Finance Companies

The amount of data created online last year would fill 57.5 billion 32-GB Apple iPads, which, stacked up, would form a mountain 25 times higher than Mount Fuji.

This “big data” presents opportunity ― and obstacles ― for auto financiers, said Don Goin, chief information officer at Capital One Auto Finance, during a technology presentation at the recent Auto Finance Summit.

Big data refers to large, complex data sets that have become awkward or impractical to manipulate with traditional database-management tools. Amazon.com’s book recommendations or Google’s ability to correct spelling and offer alternate search-term ideas are driven by big data.

“The growth of data over the last couple years is absolutely staggering,” Goin said.

For instance, Twitter users now send a billion tweets every two and half days. By comparison, it took more than three years for the social media site’s first billion Tweets to be sent.

Last year, 1.8 zettabytes ― the equivalent of 1.8 billion terabytes ― of data were created. To put that statistic in perspective, Goin said that 1.8 ZB of data is equivalent to 200 billion high-definition movies, each of which is 120 minutes long. It would take one person 47 million years of round-the-clock viewing to watch them all.

With such massive amounts of information, lenders must first scout out data-warehousing options. Goin mentioned a handful of open-source systems ― including Hadoop (www.hadoop.apache.org), Mahout (www.mahout.apache.org), and Hive (www.hive.apache.org) ― which can capture, manage, analyze and act on big data sets.

The primary challenge for lenders involves deciding how to most effectively harness the information being stored. “Assemble a nice repository of data, then spend time thinking about crunchy questions,” he said. “Then structure the plan around those.”

Marketing, customer acquisition and retention, and fraud prevention are all potential opportunities that auto financiers can leverage, he said.

Some of Goin’s ideas:

• Correlate how many people check in, using a service like Foursquare.com, at dealerships, or how many “like” a certain dealer on Facebook, then use proximity marketing to attract customers.

• Analyze how satisfied customers are with specific services, as compared with the customer experience overall. “When customers get loans at Capital One, we want them to think it’s the best experience they’ve ever had,” he said. “We want it to be really easy for them to do business with us.”

• Identify trends that correlate to fraud. Goin mentioned an effort in which credit bureau Equifax (www.equifax.com) applied big data to determine that scammers who create fake credit cards have different buying habits than typical cardholders.

Other possible uses for big data include cross-product offers, historical pattern analysis, and social media correlations, Goin said.

“This is more about exploration at this point, trying to discover and understand where this big data is going to take us,” he said.

So far, mostly credit card companies have made big data investments. American Express (www.americanexpress.com) uses the technology to increase customer loyalty through incentives. For instance, if a customer checks in at a certain restaurant, AmEx may instantly offer the cardholder a discount.

Barclays (www.barclaysus.com) uses big data to identify retailer-specific trends, deepen customer profiles, and identify and predict spending patterns.

“For all the hype cycles we see in technology, I think this is one [trend] that we should pay attention to,” Goin said.

―Marcie Belles