Big Data in Financial Services

Print Friendly, PDF & Email

In this special guest feature, David Friend, co-founder & CEO of Wasabi Technologies, takes a look at the big data and cloud storage technology stack as it relates to the finance industry. Prior to Wasabi, David co-founded Carbonite, one of the world’s leading cloud backup companies. A successful tech entrepreneur for more than 30 years, David got his start at ARP Instruments, a manufacturer of synthesizers for rock bands, where he worked with leading musicians of the day like Stevie Wonder, Pete Townsend of The Who, and Led Zeppelin. David has also co-founded five other companies including Computer Pictures Corporation – an early player in computer graphics, Pilot Software – a company that pioneered multidimensional databases for crunching large amounts of customer data for major retail companies, Faxnet – which became the world’s largest provider of fax-to-email services, as well as Sonexis – a VoIP conferencing company. David graduated from Yale and attended the Princeton University Graduate School of Engineering.

The financial services industry is highly competitive, with products fighting for the smallest differentiation to make an impact in the market. Understanding and predicting consumer behavior is key, forcing companies to ask questions like, what offers will attract new depositors? What impact will fees have on churn? What will be the effect of new branch locations on customer loyalty? What is the likelihood that the use of one banking product will lead to the use of another? And so on.

The answers to many of these questions lie hidden in the company’s historical data. The more detailed data one can gather and analyze, the more accurately one can predict the future behavior of consumers.

The first industry to apply this method of analyzing started long before “big data” was called “big data.”  Insurance agencies were among the first to make heavy use of statisticians and actuaries, especially when it comes to policies that depend on playing the odds. Take life insurance for example. If you live past your statistically-likely lifetime, you “lose” and the insurance company “wins” because they get to collect premiums from you for a longer period of time, and an insurance company that incorrectly forecasts probabilities faces the risk of potentially going out of business. Before computers, the insurance industry had to be content with analyzing by hand very crude publicly available data on average lifespans, incidences of fires and floods, and so on. These days, considering the mountains of highly detailed information available, it is possible for insurers to price products so that they align well with the real risks. In theory, having “big data” benefits both consumers and companies. By reducing the likelihood of the insurance company placing a bad bet, they can price their products to reflect a reduced level of risk.

In this day and age, this approach has proved to be just as effective in the financial industry.  Take fraud detection in the credit card industry, for example. When my wife goes to the ATM, she almost always takes out the same amount of money from machines near our house. If someone steals her ATM card and tries to use it, it’s likely that the thief will try to withdraw a different amount of money from an ATM in a different neighborhood, thus setting off alarms with the card’s fraud detection system. What makes this kind of service practical is an understanding of the normal behaviors of millions of customers, and that means accumulating a lot of detailed transactional data over a very long period of time. Not only does this reduce the credit card issuer’s losses, but it filters back to the consumer in terms of lower fees or more perks.

It’s also proved effective in marketing. Financial services companies have a broad range of products, from savings accounts, investments, credit cards, insurance, consumer lending, mortgages, business loans, educational loans, and more. And, by using big data, they can market these services to customers who might actually be interested in them! To use simplistic examples, a banking customer in his 60s who already owns a house probably isn’t a good prospect for a mortgage or a college loan. A customer who travels a lot (you can tell because they use ATMs in other cities – you don’t have to scrape Facebook information) might be a good prospect for a travel-related credit card or a airline or hotel loyalty program. A teenage customer might be a good prospect for a college loan, and so forth. The more you know about the customer’s behavior, the more likely you are to market something to a receptive audience.

Storage for such “big data” applications, however, needs to be both cost effective and fast given its sheer size and volume. If storage is too expensive, the costs may outweigh the potential gains. If it’s too slow, the analytics will take too long to crunch through to produce useful results. This is where the type of data storage comes into play. If you think back to the pre-computer insurance company days, it was impossible to differentiate between high-risk and low risk customers. With the abundance of information available today, and the computing power to crunch through it all, companies can slice finer and finer segmentations of their market, thereby more accurately aligning prices with risks. That’s why as price drops, there becomes an exponential relationship between the cost of storing data and the value and benefit derived. You wouldn’t store twice as much data if the cost was cut in half – you would increase it to 4x. If you think of a pyramid of benefits where the big ones at the top can be addressed with a small amount of highly aggregated data and the bottom of the pyramid where there are hundreds and thousands of ways that data can make small improvements, it’s clear that every time you drop the price of storage, a whole raft of applications become economically feasible.

Some companies are working to cut the cost of cloud storage by 80% over incumbent cloud storage solutions like Amazon S3. As these so-called “Cloud 2.0” services gain acceptance, there will be an enormous number of new benefits for the financial services industry. New, unanticipated, applications will pop up that would be unthinkable at the old cloud storage prices.


Sign up for the free insideBIGDATA newsletter.

Speak Your Mind