Keep calm and tweet on: Creditors have been sizing you up online for years

Last week, reports surfaced that a number of startups have begun to factor in social media data, like who your friends are on Facebook, to determine your credit worthiness. The initial report, from CNN, made big news and spread around to countless publications, causing quite a stir on the social network. 

Thing is, this story is not actually new. The idea that social-media connections and online activity can affect a person’s ability to get a loan has been around for years. Most of us just haven’t been paying attention.

First, social media data does not affect your official credit score – as in, the one you receive when you request a credit report. The factors involved in figuring out that score, which is known as a FICO score, are highly, highly regulated under an extensive law known as the Fair Credit Reporting Act (pdf), or FCRA. Social-media connections and online activity may be used for determining credit worthiness, but they are not part of your FICO score just yet.

Second, CNN buried a key fact deep down in its article: Two of the startups the report mentions, Lenndo and Kreditech, are not U.S. companies, and do not do much (if any) business in the U.S. In fact, as CNN mentions, Lenndo only provides loans in Columbia, the Philippines, and Mexico.

OK, so now that we have what’s not happening, let’s dig into what is. Long story short: The long-respected FICO credit score is no longer the only factor that matters. And credit worthiness predictions are increasingly taking whom you know and what you say on social media into consideration.

For example, another credit-assessment startup, Neo, investigates whether a person has the job they say they have by looking at LinkedIn contacts, according to The Economist. Neo has also, as Time reports, begun investigating whether people who make racist statements online are less credit-worthy. Given that companies like Kreditech use some 8,000 data points in determining lending risk, we can safely assume there are a slew of unknown activities that we all do online that go into determining whether we should be allowed to get a mortgage or a car loan. 

In other words, credit worthiness is one area where what you say and do online can have major real-life consequences.

This brings us to the most problematic part of this emerging system of risk assessment: Most of us have no idea what is being used against us. With FICO and other traditional credit ranking systems, the factors are fairly straight forward: Pay your bills on time, borrow a number of different loans and pay them all back, and keep low balances on the credit cards you have available. It’s more complicated than that, but not much.

As social factors come into it, however, it becomes virtually impossible to know how to behave properly. And there is little incentive for the companies calculating our risk assessments to divulge that information; if they tell us what not to do, they will have a more difficult time figuring out who is really a bad borrower.

Another reason that we don’t know what to do and not to do on social media is because the people calculating risk don’t know either. As Peter Fader of the University of Pennsylvania’s Wharton Customer Service Initiative told Knowledge@Wharton in February, “”It’s going to take years to understand what measures are truly valid. It’s the Wild West … like the early days of FICO.”

In other words, our credit worthiness is now being subjected to countless experiments, and that’s a problem – one that we likely can’t avoid, or stop, until after things start to go terribly wrong.

In June of last year, for example, credit agency firm Experian launched a product called Extended View, which factors “non-credit bureau data assets” that can help lenders provide loans to the estimated 64 million people in the U.S. with little or no credit history. And ZeistFinance, founded by former Google CIO Douglas Merrill, uses “Google-style machine learning” to analyze “thousands of potential credit variables – everything from financial information to technology usage.” ZeistFinance claims its credit assessment model “leads to increased credit availability for borrowers and higher repayment rates for lenders.”

In short, it’s currently impossible to give any specific advice about what not to say, or who not to “friend” or “follow” online, to help you get a loan – the factors involved in that are still evolving. So for now, the best thing you can do is keep your social circles tight, your controversial opinions to yourself, and always pay your bills on time. Everything else is up in the air.