How Banks Are Cashing In on Your Social Media and More

When thinking about the banking industry, one would not often associate a tweet with subliminal messaging. But how far are banks willing to go to make an extra buck?

The data stream being pooled each second is astounding, and we help it flourish subconsciously. Data is being collected with each like on Facebook (META -0.99%), each tweet on Twitter (TWTR), and every photo posted on Instagram. And the banks potentially have found their golden ticket! By using that information we freely share with the world, the banks could soon have influence on more than our money.

Made you curious? Check out today's financials episode with Motley Fool analysts John Maxfield and Kristine Harjes.

A full transcript follows the video.

Kristine Harjes: Can big data change banking as we know it? This is Industry Focus.

[INTRO]

Welcome to Industry Focus: Financials edition. I'm your host, Kristine Harjes, and as usual I've got The Motley Fool's senior banking specialist, John Maxfield, on the line. We have got big data on our minds this week. We produce more data than we even know what to do with nowadays and this rate is speeding up exponentially.

When you look at some of the stats and it's absolutely insane just how much information we're producing. YouTube users upload 72 hours of new video every single minute of the day. Twitter users tweet 277,000 times in a minute. Facebook users share something like 2.5 million pieces of content every single minute. It's insane.

So, the amount of data that every single one of us produces, clearly, is just mind-boggling. So, the question becomes: what do we do with it all? Businesses of all types are giving thought to this question and trying to stay ahead of the big data game, but since this is Industry Focus Financials after all, I bet you savvy, savvy listeners know exactly where I'm going with this.

We're going to discuss today what big data can do for the banking industry and just how much of an impact new initiatives are going to have. However, before we even get to the financial side of the topic let's just start high level. John, what even is "big data"? What makes it different from regular-sized data that businesses have been using to drive decisions for years and years?

John Maxfield: That's a great question. If you listen to a lot of people in the fintech industry who don't bring a background in financial services or banking to the practice, one of the things you'll pick up is that there's this perception that now these companies have these opportunities to use all this data, whereas before there wasn't. Well, there's always been an opportunity to have data -- to use data to make decisions in your business.

The difference now -- to your point -- is both quantity of data that is being produced, the multiple sources through which it's being produced, the power of econometric models to analyze that data before our regression models allowed you to look at and isolate, say, half a dozen variables. Whereas the models statisticians have come up with today allow companies to analyze 100 variables, 150 variables, 200 variables. So, a very large amount of data.

The final thing is that you have computing power now that is strong enough to take all of these inputs, throw it into a model and then organize it in a way that banks, or any type of company -- Amazon (AMZN -1.53%), Netflix (NFLX -4.35%), Pandora (P) -- can use it to make predictions about what their customers will want. It's those factors that make big data "big data," as opposed to what used to be just "data."

Harjes: So, the opportunity, clearly, is just huge. There are a number of ways that different banks are leveraging all this information to try and improve their businesses. When I first gave the topic some thought and we decided that this was going to be our episode today the first thing that came to my mind was whether or not big data could help banks analyze credit risk.

You would think that having a more complete understanding of a person's financial situation from all the different data that they produce -- maybe even something like personality aspects implied from social media behavior. Could this help lenders better assess risk when making loans? John, what do you think? Is this an area with promise?

Maxfield: Well, that's the theory. That's certainly the theory, and there are businesses that are being built upon that theory. ZestFinance is a perfect example. LendUp is another example of this. So, if you look at a bank, you go to a bank, you get a mortgage; there's only a handful of variables that they're going to look at to determine whether you're a worthy credit risk or not.

They're going to look at your income statements, your balance sheet, how long you've been employed, how old you are; all these various things, to decide whether or not they think that if they loan you money that you'll pay it back. These start-up companies are saying "There is all of this other data that we think is correlated with a person's credit worthiness and some of this data can be extracted from your interactions on social networks like Facebook, Twitter, LinkedIn (LNKD.DL), and whatever else."

These companies are building models based entirely on that. So, you have a company like LendUp; they're a data-driven company, and they're building credit models based upon how your behaviors on the social networks, but they're limiting the amount of loans -- the size of the loan -- that they're making to an individual customer to $250. So, while it's clear that they think there's an opportunity here, they're certainly not going whole hog and lending on a house based simply on the fact that you've posted seven pictures of a calculator over the past week on your Instagram account.

Harjes: I wonder if you could ever come to manipulate that. I'm going to be taking out a loan soon. Let me tweet lots of math things.

Maxfield: Yeah, or just talk about how responsible you are all the time. Perhaps that would work. We should try it.

Harjes: Just busy getting my normal dental checkup.

Maxfield: Just paying my bills on time -- a little bit early again this month.

Harjes: Be right back. So, you think those models should be more profitable. What about -- let's take this into the marketing side of things. Can you talk a little bit about how we might be able to leverage this sort of data to better market to people?

Maxfield: Right. Marketing is really the big opportunity with banks. You look at -- let me just talk more generally. The theory over the past 80 years has been that a financial company, the way they will maximize their profitability is to sell a whole bunch of products -- different financial products -- over a single umbrella.

So, insurance, checking accounts, savings accounts, CDs, investment products, asset management stuff; all these different things. The theory is that if you can get people into a checking account, then you can sell them a mortgage, you can sell them a credit card, you could sell them a savings account and another checking account, maybe a savings account for their kids to go to college. The more of those products you build up, the more profitable that bank will be.

Well, one of the issues is that -- what they've found is that in these huge organizations like a Wells Fargo that has over 90 individual businesses is that the data that is created in each of these businesses was not being shared with the other divisions. So, just merely by putting all of that data together and creating a composite sketch of your customer, you can know all these different things, all these different financial products that they have, and you could also know the financial products that they don't have.

Then you can look at -- particularly if you're that customer's primary bank -- you can go into their transaction history in their checking account and determine what their needs are, and whether you have products that you offer that could help them with those needs. Then you could send that information to one of your call centers that's making outbound marketing calls and target those people really, really carefully and increase what's known as your "conversion rate."

Just to give you an example of how effective this is, US Bancorp (USB 1.06%) -- which I've talked about many times in the past, one of the best run banks in the country -- they have used this to increase what's known as their "lead conversion" by a factor of two. So, let's say that US Bancorp draws up a list of people to call to try to sell mortgages to. Let's say there's 100 people that they call. Previously, under the system that they used to draw up this list, maybe 10% of those people would agree to a mortgage.

Now, with a more narrowly tailored -- using this information to draw up a more narrowly tailored list -- they're getting 20% of those people to sign up for a mortgage. So, it's a hug opportunity for banks.

Harjes: Yeah, that's insane -- doubling. It makes sense if you're giving people what they want. People are hearing fewer offers that they don't want, and more that they do. So, theoretically you should be improving the customer experience here too, and hopefully improving your attention because of that.

Maxfield: That's exactly right. Another really interesting thing that these banks and financial services companies are going after is something called "card-linked marketing." Let's say you have a credit card and you shop at lululemon and Starbucks all the time. Your bank can go in and see that, and then they can give you -- they can shoot you on your mobile device -- theoretically, it's still in the process of developing all this -- they can then shoot you individualized offers for the merchants that you frequent most often.

So, let's say you got to Starbucks all the time. Your bank could see that and they could say "We have all these customers that go to Starbucks all the time. Why don't we approach Starbucks and try to get some sort of rewards deal going with them to where if you buy five lattes on your BankAmericard every week you'll get a sixth one for free."

So, they can tailor that really specifically to you. The thought process in doing so -- from the bank's perspective -- is if they can more narrowly tailor the rewards programs for each and every one of their customers, that will increase the amount of credit card transactions that those customers undergo. By doing that you increase the amount of money that the bank earns both in interchange fees and if the customers carry balances and interest income.

Harjes: Do you think geographically based data could also come into picture here?

Maxfield: That's a big part of the conversations. They're saying "They have these phones," you can opt in, you'll download an app, and they'll generally ask you 'can we use your geolocation data to send you offers? Banks are talking about combining that with this card linked marketing. Let's say you're walking around in Old Town Alexandria, you're walking by that Chipotle (CMG 0.38%) and your bank knows that you've charged on average three times. You've bought lunch at Chipotle three times a month.

If it knows that you're walking by that Chipotle it can then shoot you an offer to that Chipotle and if it's around lunchtime and it knows you like it, it can shoot you that offer and that could maybe steer you into Chipotle to then make a purchase. So, this is something that's still in its -- at the beginning, but it's something that is at the forefront of these banks minds.

Harjes: That's crazy. Do you think there's any sort of opportunity here for safety? If your credit card is getting stolen, or something like that. Monitoring your personal trends to see if anything looks potentially fraudulent?

Maxfield: There is. That's exactly right. Let's say that you are -- this is one of the examples that I came across. Let's say you're at a conference in New York. You live in Washington, D.C., you're at a conference in New York City, and you're tweeting "This conference is so awesome! New York City is so sweet! I love the Big Apple!" Let's say you have a secondary credit card that you don't use very often, it just doesn't have a lot of transaction data to determine whether a specific transaction is out of the ordinary for you or not.

If that transaction occurs in Los Angeles at the exact same time that you're tweeting about being in New York -- if the bank was following, or had some sort of model that could follow your activity on your social network, it could say "We know that Kristine is in New York right now at this healthcare financial conference, but her credit card is being used in Los Angeles. So, we know for sure that's not Kristine using her credit card. Let's shut that off and then contact Kristine and go through the process there."

The other side of that -- to the point you made earlier about the customer experience -- lets' say you're traveling in Europe and you didn't tell your bank ahead of time that you're going to Europe and you had never been to Europe before. So, your credit card company starts seeing all these unusual charges in Europe on your card. One could easily think your card had been stolen and was being used abroad when you're just staying at home.

But if you're over there tweeting, or posting on Facebook, that you are in Europe, and your credit card company knows that, it will then have no incentive, and will not be as concerned about shutting your credit card off and leaving you with no money abroad. That's known as a "false positive," and that's another great way that banks can look at all this data that's created external to the organization, and that relates to the customers to improve the customer experience.

Harjes: Wow. So, there's no question that big data will come to shape the future of the banking industry. It remains to be seen just how much of an impact it will have. I, for one, can absolutely imagine in the next decade walking around some new city and having my phone buzz with a location based offer from that Chipotle that I'm passing, or an instantaneous shutdown of some stolen credit card like you mentioned.

The very first time it's fraudulently used, or any of the other scenarios that we've discussed today. I kind of get the impression that some of the industry landscape will be fundamentally different in ways that we haven't even imagined.

John, thanks so much for sharing your insights with us today. Folks listening, thanks so much for tuning in. Until next time, have a great one everybody.

As always people on the program may have interests in the stocks they talk about, and The Motley Fool may have formal recommendations for or against. So, don't buy or sell stocks based solely on what you hear.