Brennan Knotts, Product Lead at MX, introduced FinSmart in our most recent webinar, which you can watch here. (Brennan's section starts at 26:15.)
We've also written out a slightly condensed and edited transcript below:
The MX Platform
At MX we build money management tools that empower the end user, which ultimately drives profitability for banks and credit unions. We’ve done this in the past with some traditional tools that you might recognize from the PFM era. But we’ve also built some digital money management tools that are completely new to the age of advanced analytics. What’s more, we’ve done it all on top of a platform that has the best external aggregation and industry-leading categorization accuracy. We take a lot of pride in that.
Today we’re going to see the fruits of building that powerful platform as we walk through a specific example of personalization. We’ll see some of the current functionality that MX offers to our clients, how we deliver it, and what it ultimately means for the end user.
We’re going to walk through a feature that we call FinSmart. FinSmart is all about personalization. How do we take all this information we have about the end user and deliver a message that keeps their interests in mind?
Let’s start with a user who we’ll call Ashley. Ashley logs into her mobile banking app and checks her balances. You can see that she can also deposit checks, pay her bills, and transfer money.
In addition, if you look at the bottom of the screen here you can see that she has a notification, a FinSmart notification.
She scrolls down and reads this high rate warning: “Your rate is 19% higher than your peers. That’s an extra $438 a year in interest.”
Now there’s a lot going on here. Let’s look at the first bit of data: “Your BankAmericard rate is 21.24%.” This is very specific to Ashley, and that’s important. FinSmart is a marketing tool, and the end game here for the financial institution is to drive new account holders. But it shouldn’t look like an ad.
Right out of the gate, Ashley can see that this message is specific to her because she knows she has a BankAmericard credit card. She can see right out of the gate that this message is specific to her.
So what? Why should Ashley care about this message? Because Epic Bank can save her $438 a year. Again, this part of the message is very specific to Ashley at this point in time. Which raises another question: Why is Ashley receiving this message? We’re not showing this message to all users since that wouldn’t be very targeted. What about Ashley’s situation allows us to know that this is the right message for Ashley at this time?
There are a few things going on in the background.
First, we’re listening for certain data points, and in this case we know that Ashley has an externally held credit card. She’s got this BankAmericard credit card, but she’s also banking at Epic Bank. And Epic Bank might even be her primary bank given that she’s opening their mobile app.
Second, we also know that this credit card has a higher interest rate than the rates that Epic Bank is currently offering.
Third, and perhaps mostly importantly, we know that Ashley is carrying a balance on this credit card from month to month. If users pay off their card every month, they probably don’t care what their interest rate is, and so with FinSmart they wouldn’t see this notification.
How We Get the Data
We’ve just looked at the first touchpoint with Ashley. We’ve let her know that we have this specific message for her. Now she can click into it and learn a little bit more about what is her scenario.
Again, this is designed and optimized not to look at all like an advertising experience. The idea with personalization here is to deliver a new feature to the end user. So Ashley can see here why you are delivering this message to her and how are you arriving at this idea that you can save her $438.
Let’s look at these different data points, starting with the current APR on the BankAmericard credit card. Where is MX getting this? The majority of the time we’re actually getting this directly through an aggregation feed. So Ashley doesn’t have to do anything. This rate is automatically pulled in. We’re automatically using this to tell Ashley something she may not know about her financial situation.
This next data point is the estimated interest cost. This is key in understanding how much money Ashley could save. We’re getting this from the transaction data. We’re able to look at Ashley’s past transaction history and see that every month she’s getting hit with a finance charge on her BankAmericard credit card, which tells us she’s carrying a balance. And the amount of that finance charge allows us to extrapolate and understand how much is she actually paying every year to carry a balance on that credit card.
Then we get to these next portion, which is where Epic Bank gets involved and tells the system a little bit about their offering so that Ashley can understand how Epic Bank can help her save money. In this case, Epic Bank tells us that their interest rate is 7%.
With these three data points we know how much money Ashley could potentially save if she transferred that balance to this lower rate credit card. And of course down at the bottom there, we have the Apply Now button. So what is the call to action? Ashley is seeing that there is an opportunity for her to save money here and she cares about that. $438 a year means a lot to her. What does she need to do next? This will lead her to the online application page. This is what the experience is like for an end user like Ashley.
Why the MX Platform Matters
What makes this possible? How are we able to deliver this today to the end user, and how could you potentially do this within your own organizations? The first key component here is the external account aggregation. This is what allows us to understand Ashley’s full financial picture. It allows us to understand that she has that BankAmericard credit card. Not only that, but we have specific data points about that, like that interest rate. Without that information, we really can’t tell her how Epic Bank might be able to save her money.
Another part here is what we at MX call the 3 C’s. The 3 C’s stand for cleansing, categorization, and classification. You might have all this data — it could even include external data — but you can’t get a lot of value out of it if it’s not clean. It might be inaccurate. You might not be able to understand that this is a transaction at a grocery store, while this is spending at a gas station. Those nuances matter if you’re going to build a personalized experience that the user is going to want to take action on. So we have to get all of that right and in this specific instance the transaction data we care most about is that finance charge. We see where Ashley is getting hit with different charges and what those mean about her past history.
The third part of this is dynamic targeting. This is what takes an experience that is lackluster and gives it a little wow. In this case, I mentioned that Ashley receives this notification and no one else does. And the timing is also critical. If Ashley paid off her credit card yesterday we don’t want to show her this message. So the system needs to automatically recognize that scenario and drop Ashley from the segment of users it will show this message to. Or if Ashley charges up her credit card a ton yesterday — maybe she bought a couch — then her real-world savings could possibly be a lot more than what we were presenting yesterday. In this case FinSmart knows how to update that information in real time. This happens every time Ashley comes in and looks at her finances. In fact, even when she doesn’t open up the app we’re still updating that information in the background.
Another key component of this is the dynamic messaging. Even if you get really good at understanding the user and pushing them a message at the right time, they’re going to zone out if you don’t make that message specific to them. They’re gonna ignore it the same way they ignore banner ads in the sidebar of their favorite news site. It’s not enough to say, “Yay, it’s Friday!” or “Happy Birthday!” since those are a little personalized, but not enough to catch your attention. “Yay, it’s Friday!” is not specific to Ashley. She’s going to tune that out almost immediately and move on. So it’s very important in the way we deliver this message that she gets right away that this message is specific to her.
Simple to Launch
We’ve looked at what the experience is like for the end user, but let’s be honest. If you’re going to take advantage of this functionality at your organization it needs to be easy for your different team members to use. You need tools that make it easy to configure the personalization with your own institution’s information, but you also need to track it easily. How are you justifying the time spent here, or the money invested in delivering personalization? If you can’t easily track it, you can’t justify it. It comes in danger of never getting off the ground or being dropped eventually.
Behind the scenes, there’s a data repository and a lot of hardware that processes all of the transaction information that’s coming in. We’re talking about a big data marketing platform that’s concerned with governance capabilities, security, data preparation, data maintenance, and more. And there’s Forrester research that says if you build a big data marketing platform in-house, you’re looking to spend between $25 and $50 million dollars just to get it off the ground — not to mention ongoing costs. You can implement FinSmart with a small fraction of that cost.
Now let’s look at the experience we’ve built for the marketers at your institution. What are they going to see when they’re using this functionality?
We call this tool Insight & Target. Insight lets you quickly track internal and external data and Target lets you create campaigns within minutes.
Target is where the FinSmart campaigns are configured.
You can see the low interest credit card savings campaign here.
Let's look at how to set up this campaign.
First, you select a product name. Second, you choose the URL you want them to visit to apply for the credit card you're offering. Third, you enter your interest rate. Fourth, you add a little eye candy — an image of that credit card to associate in a user's mind an image of what we’re talking about. Fifth, you preview and launch the campaign.
Just like that, people like Ashley will start seeing a hyperpersonalized notification about switching her credit card to Epic Bank.
Tracking is pulled in automatically. You’re going to be able to review this FinSmart campaign that you launched, and you’re going to be able to see how many users actually fit this campaign. How many have already had an opportunity to view it, and how many have actually clicked on it and had an opportunity to take action and hopefully apply for a credit card.
Again, the whole concept here is to make this super easy to use, and make this automatic. Once you set this up, no one needs to come in and manage the segment, or update the segment of users who are seeing this. All of that is happening automatically in the background, and it’s always on.
So far we've only focused on the credit card example, but there are many more.
You can look at mortgage scenarios where users have a mortgage with you, but may now qualify for a better rate since they've improved their credit score. Or maybe you can see that they have a mortgage with a competitor and again, maybe for the same reasons, maybe they have a higher rate than what they could potentially get today.
You can do the same thing with debt consolidation. Through external aggregation you can see users who have loads of debt and ask how can you help them consolidate that. Personalization makes it possible to do that without having to involve a lot of human resources. The software is going to handle that directly for you.
But even getting beyond just looking at the products and services you can offer the user, why should they engage with your platform beyond that? You can tell them things about their spending. In terms of providing personalization, at MX we have a philosophy where we don’t want to come in and lecture the end user. We don't say, "Stop spending money at coffee shops," for example. Instead we want to point out relevant data. We might say, "Your spending in this category is more than your peers’ spending in this category." We raise these interesting points for the end user to ponder so they can feel like they have a better pulse on their financial situation without having to sit down and do a deep dive with an Excel spreadsheet. The software does it for them automatically.
In conclusion, why is personalization really important right now? It really comes down to how you’re going to grow your financial institution. In the past, one great method to grow your financial institution was through acquisitions, especially when there were some distressed assets available at a low price you could grow through acquisition. In the past you could also grow through really aggressive traditional marketing — really pouring a lot of money into direct marketing campaigns or TV advertising. This didn’t necessarily yield the highest percentage conversions and could be really expensive for not really knowing what your results were going to be.
Today the traditional advertising model feels a little broken, which has led other industries and the financial industry to a more data-driven approach. And that’s what we’re talking about here with FinSmart and personalization. How can you use the data to retain more users? How can you use the data to encourage them to use more of your products and services? And also to win more of those held-away accounts that they’re currently banking with competitors.
So I’ll end on a quote that I think sums up a lot of what we’ve been talking about today in terms of mass personalization: “Mass personalization tied to a permission-based money movement is going to change this industry so dramatically that we will not recognize it within the next decade.” This comes from Bradley Leimer, Head of Innovation at Santander. It's one man’s opinion, but it also happens to be MX’s opinion. If what I just showed you is possible today, think of what’s possible within the next year, or two years. Let alone the next decade. I think it’s truly going to be remarkable and it’s going to completely change.
Picture this scenario: Users log into online banking to see what their balances are, and you tell them the action they can take to change their finances for the better. That's what the future holds.