One lender uses sophisticated algorithms--and any online dirt it can find about your company--to determine credit risk.
The amount of stored digital information that companies, researchers, and others can mine is growing exponentially and is nearly impossible to fathom. In the book Big Data: A Revolution That Will Change How We Live, Work and Think authors Viktor Mayer-Schönberger and Kenneth Cukier maintain that in 2013 the world's collection of data amounts to around 1,200 exabytes. "If it were placed on CD-ROMs and stacked up, they would stretch to the moon in five separate piles," say the authors. "The amount of stored information grows four times faster than the world economy, while the processing power of computers grows nine times faster."
What does this have to do with you and your company? A great deal if you want to get a loan.
The Big Data Back Story
Every digital act you make is stored--think Tweets or Facebook likes or what you buy on Amazon or what you search for on Google not to mention your online bank account, your credit card transactions, and even where you are based on your cellular transmissions. An interested entity with the resources to do so can create profiles that pretty accurately portray who you are as a person.
For small businesses it's the same thing. Your digital cash flow, whether or not you've had property foreclosed upon, what others are saying about your company online and more can be tracked, aggregated, and analyzed so that conclusions can be drawn.
Mayer-Schönberger and Cukier talk about how the ability to analyze billions of data points instantly has huge implications for society. There are cool things like Google's ability to show in real time where the flu is breaking out based on the terms people use when they search the Internet. But big data also ushers forth not-so-great implications, such as the fact that privacy is pretty much dead in the water.
Prediction is another side of big data, one that's either injurious or invaluable, depending on your perspective.
"[A]lgorithms will predict the likelihood that one will get a heart attack (and pay more for health insurance), default on a mortgage (and be denied a loan), or commit a crime (and perhaps get arrested in advance)," they write.
Getting a Small Business Loan
"Minority Report" and "Person of Interest" foreshadowings aside, one company says it's using big data to help small businesses.
If you own a restaurant, auto repair shop, nail salon, dentist office, or any kind of Main Street small business, borrowing money from a bank can be tough. After the recession many of them largely stopped issuing small business loans so if you need money to buy equipment or inventory it means either taking out a personal loan based on your FICO score or using a credit card.
There's another option, and one that's growing in popularity with both business owners who need a loan as well as big shot VCs who want in on the action.
New York-based start-up On Deck looks at a company's digital footprint--things like bank transactions, public records, and social data such as the kinds and frequency of reviews on sites such as Google Places, FourSquare, or Citysearch--to figure out how much risk is involved in issuing a loan.
According to On Deck CEO Noah Breslow, his 6-year-old company has delivered around $500 million in capital to thousands of Main Street businesses across hundreds of industry verticals.
"We've built a platform that aggregates all that data so in some cases we integrate with different providers of data and in other cases we go out and actually harvest that data itself from the Internet because it's publicly out there but you need to extract it, aggregate it, and get it into usable form," he says. "Then when a business owner applies, we collect as much data as possible on them from all these different sources and use that to make a very efficient lending decision usually within minutes after the customer applies."
Such efficiency is one thing that led Google Ventures to invest $17 million in On Deck last month along with PayPal co-founder Peter Thiel as well as previous investors including SAP Ventures, RRE Ventures, and others that altogether have thrown $93 million into the company.
According to Google Ventures General Partner Karim Faris, what On Deck is doing is quite different from how traditional banks underwrite loans.
"It's unique in both the data signals and data analyses that they take in and also really more mundane stuff like the process they use to underwrite loans," he says. "The ability to get it done within days versus weeks or months is actually no small feat and a significant benefit to the business. If you can get a loan within a few days versus months that could be the difference between life and death for a business plan."
But Does It Work?
The numbers say so: Breslow says On Deck has achieved 117 percent compound annual revenue growth since 2007, will double its employee head count to 300 by the end of year and is opening branch offices nationwide.
Breslow says On Deck's credit scoring algorithms are on their fourth generation. Part of the magic of its intellectual property involves how it is able to factor in the intricacies of individual verticals.
"We [are] underwriting 700 different industries that we have in our loan portfolio so we have to understand how a doctor is different than a restaurant and how that's different than a plumber or how that's different than a light manufacturer. They're all different kinds of businesses with different credit profiles, different cash flow, different margins," he says.
Breslow says On Deck loans are typically three to 18 months in length, require at least one year of business history, and are ideal for businesses that have a lot of small transactions. They're less than ideal for contractors or consultancies that don't get paid frequently or for tech start-ups or other companies that are focused on R&D.