Data is the New Gold

Of course, this is a revolutionary conclusion that I came up with before anyone else. And, of course, it is not an oversimplification.

I know. Data is not gold because, unlike gold, it is abundant and unlimited. And over the last few years, data has also become more broadly accessible. And what is abundant and accessible is not, by definition, very valuable.

I will contradict my headline further by stating that data analysis is not a solution to all our problems. Like anything new and exciting, it has been overhyped. It is an excellent tool on some applications, but definitely not all.

But there is a kind of data that is gold-like valuable in a few narrow applications. I'm talking about proprietary data, data that is only accessible to you. And I'm talking about data in the narrow field of credit risk analysis, specifically in lending to SMBs and micro-lending.

Again, quite revolutionary.

We've always known the importance of information in lending. Traditionally, banks would obtain said information directly from credit candidates during the underwriting process. Bank officials would also have access to the banks' own historical data making such loans and their own experience in the field.

But the economics are markedly different when lending to SMBs. The traditional underwriting process is time-intensive, and therefore too expensive for small loans. Just as important, most SMBs don't have all the information required by banks, such as financial statements. To top this off, SMBs rarely have the assets necessary to secure a bank loan. So yes, lending to SMBs is a different beast.

SMB loans are cash-flow-based, with recovery rates on defaulted loans around 10%. And loan underwriting is done with statistical analysis, looking at factors such as industry, location, years in business, sales, number of employees and the owner's FICO score. To do this accurately, though, you need lots and lots of data.

You can get a lot of data from public sources, but so can everyone else, so the only differentiation is getting, storing and analyzing your own data. For an SMB-focused lender, that takes time, plenty of mistakes and lots of capital, which means a costly education.

There's a catch-22 here: the more data you get, the better at lending you are likely to get. To get more data, you have to deploy more capital and do so faster, increasing the likelihood of expensive mistakes.

Is there a faster way?

Yes.

Provide services that churn out relevant data, like payment processing. This is why new payment processors such as Square, Stripe and Paypal have been very active and profitable in the SMB lending space. They generate TONS of data from their payment processing and use that to offer financing to their clients, which further enriches their database.

Terrific, right? We have to set up a payment processor and then start lending!

Well, yes, but that is incredibly difficult, capital intensive and risky.

Is there another way?

I'm betting there is, in POS (point-of-sale) software.

A cloud-based POS solution can provide the same, if not more information, than a payment processor because it handles all of a business's transactions, not just credit card sales. POS solutions can offer modular ERP-like products and services for managing inventory, payroll, financial planning, cash flow management and even accounting and taxes.

A low-priced solution that allows the POS vendor access to anonymized data would be the equivalent of a gold mine. I should have used "POS Vendors are the New Gold Mines," but that title is even crappier than the one I used.

Competition is fierce in this space, but you do not need to be the leading POS vendor. Reaching a critical mass will do.

Easier said than done.

Is this a challenging, risky and long-term solution? Yes. But it is also doable, much more so than acquiring a payment processor or starting one from scratch.

Do you think I'm wrong, stupid or crazy? Hit me with an email and tell me why.