Nandita Gupta, CPO, and MX Product Leaders
U.S. Bank, Webster Bank, Morningstar, MX, Atomic
Alright, welcome. Thank you all for coming. This is a good group. I'm excited for the conversation today. I think the 45 minutes that we've got, we'll go relatively quickly. Um,
it'll be a good conversation and, look forward to the conversations that come up afterwards as well as we go through the conference. Um, so I am, uh, Nate Caldwell, I am with MX. Um, have been with MX for about 12 years, quite a while, a lifetime in some ways. Um, and I'm excited to moderate today's panel.
Um, we've got a lot of great guests here. I am gonna give everybody a chance to kind of introduce themselves. Um, but, wanted to kind of go through and, and introduce, um, and, and I'll give another thing as you introduce yourself. Sorry. Uh, we had this question kind of in there, but, um, in the list of questions, but I want to kind of start off with an icebreaker. Ee have, um, Lisa Fitz from Regions Bank. Uh, she's the VP of Digital Experience with, um, I think this is a long title and actually I probably could read it off of their storefront and onboarding products with Regions Bank.
and then Brian Francis, who's the head of account check with Informative Research. Um, and then Abbas Moosavi, who is the lead product, uh, manager for investments with MX. And so, um, this group, I've worked with everybody on this panel a good amount, um, have a lot of respect for them and look forward to the conversation. So I'm gonna give them each a chance to go through and introduce themselves. Um, one thing that I'd like to do is you kind of speak about, give me an idea of what you do in your job today and how you work with data. And then as the icebreaker tell us something interesting about yourself, whether that's, you know, work related or like something that you like to do in your spare time. Um, like what do you do outside of work as well? So start with you, Lisa, and kind of go through.
Okay. Awesome. Um, Lisa Fitz. I, uh, live outside of Pittsburgh, Pennsylvania. I worked for Regions Bank based in Birmingham, Alabama. I've been in the financial technology space for over 20 years. Um, primarily in my last 10 have been focusing on digital experiences for
small business origination and servicing. And then at Regions I encompass, all segment types and, uh, do what we call the shop and buy experiences. So I'm really focused on getting those people to the door, right, uh, validating making sure that they're, you know, getting their needs, in the onboarding journey.
And then transitioning over to my partners to service them. so data is very critical for me. Uh, just watching, who's coming to the front door, fraud monitoring, um, watching attrition, watching churn, watching people, you know, develop into primacy. What are they doing? What are they acting with? And, it's really important for me to put products and services at the forefront so that we get them onboarded and acting and engaging straight away. Um, you know, we don't wanna waste any time, so we're really pulling in many data sources and we'll talk about some of the pain points of doing so. Um, and trying to make that a little more personalized.
That's really it. And my fun fact is I walk on water. I, I truly do. Um, and, and yes I do. And, uh, I'm an avid paddle boarder. I will paddleboard anywhere. I'll do paddleboard yoga. I don't look like somebody who does, but I go out there and I give it a try. But I'll tell you, when I'm out on that paddleboard, on a crystal clear lake, I feel like that is probably as close as close to the upper beings and walking on water because it's just nothing more moving than to see the fish swimming around you on a nice, clear lake. And, uh, so that's my fun fact. I was gonna lie and come up with a great story, but I knew Abbas would call me on it, so I didn't.
That's awesome. Paddle boarding's hard. I don't know if you haven't tried it, it's a very difficult thing to stand up on a paddleboard.
Well, I Haven't tried paddleboarding yet, but, um, thank you everybody for joining. I'm Brian Francis, and I'm the head of account check with Informative Research. So account check was the world's first use of bank transaction data for verification of assets for mortgage lending. So back in 2009, account check was really created to solve the problem was how does somebody
applying for a loan share their information digitally for lending?
So throughout the growth of account check, I've been partnered with over 3000 lenders. We were instrumental in the launch of Rocket Mortgage in late 2016. We were granted, uh, government guarantee called Day One Certainty. So the use of the account check solution by a lender meant that they no longer had to carry rep and warranty on the data that they were using in closing with the loan.
So we are recently part of the Informative Research and Steward Title family. So, what we're focused on now is really broadening the adoption of the use of digital data and branching into new areas of verification of pulling in other direct source digital data. So looking at how do I verify income employment, how do I tie that together with credit data that all exists, and then looking even outside of the personal, but looking at property, looking at valuations associated with, uh, the house that I'm in or the house that I'm buying, and really put together a true holistic piece, uh, view of a person's financial capacity.So I'm not really good on these icebreaker things.
So, um, but I will say this, I am actually a truly a rocket scientist. I do have a degree in aerospace engineering, uh, from Georgia Tech. I don't claim them for football, but, uh, it is a good school to go to, have never worked in the field, but I still watch almost every SpaceX launch and can't wait for the next launch of Starship.
I don't know how I'm gonna top these two. That's pretty amazing. Uh, hi everyone again, my name is Abbas. I am the product leader for investments here at MX. Many of you are very familiar with MX 'cause you wouldn't be here if you weren't. Um, but with investments, um, we're starting to create some really, really exciting things. And at its base level, we're starting with investment aggregation to build the vision of investments and to understand how we can segment out our customers, um, and create unique experiences for them. It all starts with investment aggregation. So as we're continuing to work through, uh, investment aggregation, I will put it out there that, uh, we released the beta phase of it at the end of August. And, uh, really excited to see where it goes.
Now, a fun fact about me, I did say a fun fact this morning. I'll change it up a little bit. Uh, I am a huge adrenaline junkie. Um, I'm one of those people that hike up mountains and I'll run off it and fly down. Um, so if you're ever into it, let me know. I'll take you on the hike, but not the flight.
Awesome. Well, thanks. Um, like I said, it's a great group, excited for the conversation today. Um, and, uh, our, our main focus, I'm sure that you've read kind of the summary of what we want to talk about with data enhancement, the power of data enhancement, and different ways in which enhanced data can be used.
Um, one of the things that I think is kind of core to data enhancement is the use cases around it, right? Um, and so what data you need, um, and, uh, really depends heavily on what you're using it for. And so, um, if I could just maybe, let's start with the boss and then kind of come back to Lisa, if that works. Um, I'd like to kind of maybe have each person talk about, you know, what are the use cases that your company is looking at? Um, and then what does that mean for the types of data that, um, you need, uh, for what you're doing?
Yeah, great question. Um, when it comes down to data, I'm really trying to focus on what is the best value that we can bring to our clients and to their customers. Uh, in the end, providing value to the customers is what MXs goal is, right? And to create that type of value, you need to recognize what type of segment these customers belong in because the experiences that you're gonna be providing to these customers can vary dramatically, especially in investments.
So as we build out the investment aggregation product and, and we're trying to build out towards like a future phase, what we're really trying to do is help our clients understand those different types of segmentations. If someone has a $5 million portfolio, you should not be giving them the same experience as someone that may have, you know, $30,000 in their portfolio, right? And so as we build out this investment aggregation product and we aggregate the data and we understand that data, we wanna be able to create actionable steps on, on that data, right? So it's creating like optimized suggestions or it's creating like benchmarking data. Um, and that's what we're really, really going towards. And then I'm happy that it falls with MX values as well too.
Awesome. Brian, you wanna go next?
Yeah. So in, in the lending use case, if you think about it, the, they, what they call the, the three Cs, uh, when looking at a loan that's credit collateral and capacity credit. That's your bureaus, that's your experience. And Equifax's collateral, that's the property that you're buying capacity is what is I, as the consumer able to do? What am I able to afford in this loan? So what tells the best story about a consumer? And that's bank transaction data, but it's looking at bank transaction data maybe a little bit differently than we, than you would look at if you were say, doing a personal finance management solution.
Whereas if I'm managing my money and I want to classify my transactions, then a data classification that fits into what a Microsoft Money Quicken to talk about some of the older products that are out there that is a little bit different than what a lender might want to look at. So being able to classify data specifically for the use case was very important to account check. And it's why account check, uh, built our own classification capabilities to identify those items of interest. We didn't wanna be making the decision, but we wanted to highlight those pieces of information that needed to be looked
Things like unusual deposits, not necessarily a large deposit, but a deposit that was unusual given other deposits that this consumer might have had things like not sufficient funds. And then when you expand out, when you see the rise of companies that are providing access to payroll data, well, where does payroll data end up? Well, payroll data ends up in your bank. So being able to classify incomes, identify multiple streams of income and from one data pool accessing a consumer's bank data really tell a compelling holistic story of their capacity in the loan, um, process.
I'm curious on that, Richard, before we move on a little bit. Um, as you developed, kind of, your own classification system and you serve a lot of different clients, right? Um, do, have you found that, even lending within different clients can vary and they may find certain data elements more or less useful? Or have you found that it's more, the lending use cases is kind of down a fair way, if that makes sense?
No, That's a great question, Nate. And if you go back to the kind of the genesis of account check, where it was this complete story of a consumer driven by digital data and you sit down with a lender back in 2009 and their eyes just glazed over and said, I just need two months worth of bank statements.
So the ability and the capacity in the data is there, the ability to consume it, and for the credit providers and the rating agencies in the secondary market to. But we will continue to, to enhance and tell the same story that this is the right way to tell about a consumer's ability to pay.
Okay. Awesome. Thank you.
I have numerous use cases, but I think the one that is really the most powerful is making sure that our employees can have, um, that omni-channel view, that 360 view of the customer. So bringing that data in digitally, making it available for people who are having the human conversation is really key.
Um, currently building this, you know, uh, employee portal for the 360 view and bringing that data in a meaningful ways, um, partnering that with our, you know, next best product recommendation experience that shows up digitally and in the branch, I think will be the most powerful use case because I think at the bottom line in our bank, um, we are people driven. We are, it's our people who make the difference. And even though we have so much digital, uh, happening, that's still, you know, what our brand is known for.
So I think it's really key to make sure that those employees have that view of what's happening. And I think we talked a lot about that, you know, we use the PFM tools with MX it's like, how do you make sure they can see what a customer has engaged with, set up? What are their goals? Um, bringing that full picture together so that, that, that you have that omnichannel experience for the employees, I think is key.
Very cool. Um, and with the area that you work in, Lisa, we'd kind of talked about this a little bit, and I'm curious to dig more into, um, you've got kind of a digital side and digital experience, but then you still have kind of a branch experience as well, right? Um, 12`and you work with data kind of in both of those as people are coming in opening accounts and working with the institution. Um, how do data elements vary kind of based off of the digital path versus the real world path?
So when you think about it, we have like start anywhere, finish you anywhere. So as a user, I can start at home, finish up in the branch, and vice versa. Um, when you think about it though, when you're onboarding, you know, a user to be a customer, and most of our experiences, 70% of all the data elements you're gonna ask are the same, right? You've got the nuances by segment or by product that you're going to differ.
Um, but when, what the advantage is that, you know, we still are trying to figure out how to marry is whenever someone's having that conversation with the branch, the branch has a needs-based conversation. It's structured, it's um, you know, it's called our GreenPrint conversation. Bringing that in a digital format will be really key because that employee is getting the advantages of the cues, the nuances, the, you know, what, what's going on in that person's lives firsthand.
And you don't always get to capture that digitally or people don't wanna invest that time to give that to you digitally. Um, I remember years ago we would use the notes fields, the user to find fields, when I worked at a credit union and you would learn all kinds of things about, you know, that platform teller, you know, member service rep, they knew everything about that person. Um, birthdays, big events, right? We wanna know what, what events in your life that we can be there and be there in the moment. And it's how do we bring that conversation digitally to marry those two up, I think is, is key. So the experience could be the same, um, working towards that, but I still think that that, that personal conversation, those just ad hoc data points that are brought to that banker are still really the goal too, is bringing that into, to a more digital space.
Awesome. I love how much you focus on getting the employees the right data to serve the users. That resonates a ton with me. I love it.
Yeah, I think that the consistency, that's a great point, Lisa, is that you've got all this data about a person, but being able to enhance that data and consistently tell the same story, because as you have different people looking at data, they're going to see it through different lenses.
So it's important for the technology to be able to take the enhancements to the data, the attributions, the classifications and the categorizations, and tell a consistent story. So yeah, you may know that, uh, this person here has a birthday this month, but it's important to know that the next person that comes in, you have the same type of data about that permit.
And that's where the enhancements to data of not relying on that person to interpret it, but letting the system give you those hints, that's where you can have that much better experience for your consumers and for your associates too. 'cause they'll feel better about interacting with them.
Yeah. I wanna partner on that. You know, you get, since covid, the turnover for frontline employees has impacted us all. When you have something that's structured, it gives them the confidence as they're onboarding as an employee to have that conversation. You know, you think of someone that's 18, 19 years old and has probably spent most of their life behind a screen
To actually have somebody sitting at the desk and being able to have a conversation that you feel like you're, you're, you're really helping them. I think that's where the difference can be made.
That's awesome. Yeah, it's interesting. I think there's the normalization of data and that's one thing that we see a lot with pulling data from disparate places, but then also like the normalization of making sure that you have that, in a way that your experience and how that data is being used as consistent as well.
That's a great call out. Um, so I'm curious, maybe to jump to a slightly different, um, uh, area, with investment data. So Abbas, that's a huge focus for you. Um, and with a lot of the data we've been talking about thus far is like user data. And we can talk about retail data I think, as well as far as those retail transactions. But with investment data, I'm curious, how does that kind of vary from like what you see with retail data enhancement, um, and, and the investment space?
Yeah, great question. When you loop in investment data on top of retail, I know we're varying the two, you get a holistic overview of what that individual is, right? Like you can tell whether an individual is well allocated in their, in their retail accounts, if they're well allocated in their investment accounts. Once you loop that in and then you can start assessing that customer's profile, you can start determining their risk tolerance, you can start determining their net worth. You can start determining, um, how much cash they actually have available to.
'cause when you start getting into the really wealthy individuals, wealthy individuals, they're not touching the options that have been exercised. They're taking loans against the options that have been exercised. And from our outside, outside view, we look at that as a liability. But in their eyes they're just, they're like, no, we're, we're pretty well off right now. Right? But you don't get those insights unless you bring in investment data, right?
And as you bring in that more and more data, going back to what my first topic was about customer segmentation, now you can build products around those different types of experiences based on the segment that that customer is in, right? And once you build out those experiences, now you've developed a, you can develop a personal relationship with that individual and then you can cross sell products.
You can think about how you can expand to like their expanding network. So for example, if a personal banker developed a relationship with me, most likely they can develop a relationship with my kids, with my wife, um, with my mom and dad. Um, and so there's, there's a lot of, a lot of opportunities there as you bring in more and more investment data, um, and yeah, that everything can be derived as you bring in all that data on top of retail.
I just wanna add something. Do you find that, you know, people get pretty proprietary inside of a financial institution about their customer. You're talking about blurring the lines, right? You can blur the lines between, you know, bringing investment data in between a commercial customer from, um, someone who's got net worth, right? And, and even down into the consumer level, you know, um, you know, how, how do you think we're gonna mature? We're able to blur those lines and become more about that customer as a personal, you know, experience instead of like mine and yours. Because I know that's something, you know, throughout my career, it's like, well, that's our customer. Well, that's, no, no, that's my customer. You know, and, and how do we get people to start thinking about it? Well, no, it's the customer, right? With a special experience.
Yeah, that's a, that's a great question. Um, lemme tell you about the behavior that's going on with customers today, right? You have something that's going on right now, it's called the Great Wealth Transfer. It's when a generation is starting to transfer their wealth over to the next kin of generation. This next kin of generation have a lot of different needs. Um, and the way they develop relationships is much more different than the older generations. Unfortunately, there are a lot of financial institutions that are not adhering to those changing experiences.
It's about who's gonna get there first and who's gonna grab the market share, unfortunately, right? It's not gonna be a waiting game. It's not gonna be, oh, you know what, they're gonna stay with us because their grandparents were with us. It's not gonna happen.
If I were to acquire a massive amount of wealth from my grandparents, I can tell you right now, I'm not gonna be going with the bank that my parents are using, right? I'm gonna be more focused on financial institutions where, and this is me personally, by the way, this is not gonna be resonating with every one of you. I'm gonna go with an institution that has a hybrid model, right? Uh, because there are components of finance where I am still learning and I want to learn from that from a really good financial advisor.
But then there are some things I wanna do on my own, right? Here's a prime example. Um, look at when you pick a tax accountant, right? How many of you have ever stayed with the first tax account you ever stayed? Uh, you've ever picked? Uh, if you, if you don't go through your own taxes, like I have gone through probably five tax accountants. 'cause I feel like I was not getting the service that I needed, right? And so, uh, I hope that answered your question, Lisa.
Yeah. I think, um, the way that this conversation has started, which is great, is like focused on do you know the user? How can you serve the user? Um, and uh, using data that's provided when the user signs up, when they link accounts, helps for that. But then also the metadata that you start to say like, you know, bucket, whether it's, you know, an investment kind of group, whether it's, you know, a profile when an account's opened or even a lending profile.
I'd imagine that taking metadata and bucketing those users so you can better serve those users a big part of everybody's, uh, functionality. Um, I'm curious. I I always like to hear stories of, you know, what have you learned along the way? Um, and so, uh, Brian,I greatly appreciate the comment of we're still learning, right? Um, because I think that that is the reality of working with data is you have assumptions and you, uh, it just is really a long journey. Um, do any of you have stories about learnings along the way where you're like, Hey, this was an important moment where we learned this was important as well as potentially we learned that this was the wrong thing to do. Um, kind of take two different paths there, but I'm curious, if you have experiences in that learning journey.
So I'll go ahead and take the, um, where we learn some things. So, if you look again looking at a person's bank transaction history and what are the things you can identify looking at that information? So I would say circa 2016, we got a question from, um, I believe it was Fannie Mae that says, Hey, do you guys have the ability to identify rent payment data? And so you go and you look at the, at the, the corpus of data that you have and rent's a little bit of a challenge.
I mean, payroll, it's challenging, but it's relatively easy. 'cause you can identify, it's a direct deposit. It may have attributes for who the employer is, but how a person is paying rent, that's a little bit different. So you investigate that. You look at it, you say, Hey, alright, yes, we can identify it. Here are the patterns that we're looking for. Here are the data points that we need to be able to do that.
Well, why do you want to do that? Well, it turns out that expanding home ownership is one of the key initiatives of lenders and government agencies. And if you look at people that are looking to move into a home that are currently renting, what's their number one thing that they always pay? They might skip on an electric bill, they might skip on a utility payment, they're gonna pay their rent. And rent had never been used as a way to qualify somebody for a loan. They couldn't do it with the data that they had. You can't review physical bank statements, a person review them and find rent and do it in a way that can be certified and validated.
But by using bank transaction data, you can verify rent payment history. And two years ago, Fannie Mae came out with a program that said, if you have a certain FICO score and you make a consistent rent payment of at least $400 a month, then we will make you approve eligible for a mortgage. And the only way that that could be done was through bank transaction data.
Awesome. Very cool.
Ava, do you have something that you wanna add to that or? I have more of a personal experience.
Nothing. Nothing that deep, deep. Um, so learnings along the way. I used to be a person that used to be incredibly overwhelmed when it came down to speaking about financials, um, who to find to talk to, who to learn from. And, it honestly took a lot of initiative for me to get comfortable with the learnings along the way from when I bought my first property, um, to when I started talking about like salary, expectations to, uh, even having money in a 401K, right?
Like nowadays when I meet people, um, that are within my demographic, I ask them like, Hey, how much money do you have your 401k? Because I have no idea how much, like, am I at average? Am I above average? Am I below average? And, they would like smile at my friends would smile at me and be like, that's so weird. 'cause we would think the same exact thing.
And so, we just started kind of sharing, all those different types of financials with us. And, you know, maybe it's a Millennial newer thing that's happening right now, as you all have probably noticed. Um, speaking about financial literacy has, has been incredibly popular across all types of social media sites. And I'm talking about TikTok, Instagram, Facebook, even Reddit, right? Everyone is starting to talk about it.
Salary sharing has become increasingly popular in the last few years. That used to be a taboo, right? Um, that was a huge taboo. But now you're just, just seeing more and more people talk about it. And as people are consuming all this knowledge, in the end it comes down to are these people actually taking action on the knowledge that they're gaining from these sites, right? And they kind of look for that human interaction to help them take that first step that is incredibly overwhelming to them.
And that's what it was for me. It took, it took a friend to kind of convince me to take the first step, um, with purchasing a home. And I had regret the very next day. 'cause I was like, all right, I'm, I am cash poor now what did I do? But it was my first property and I was still learning. But along the way I just met the right people and asked the right questions. And I'm reaping in the rewards on that now.
Yeah, I would say about learnings, um, you know, looking at data from green bar paver, if anybody remembers what a green bar stacks and stacks a green bar and trying to come up with the formula that, you know, um, I think we're still very in our infancies about it, right? Yeah.
Um, I always said it, we've always been data dumb. We have all this data. It's sitting in so many silos. And I'll tell you what, the day I got a data warehouse and got, was able to bring in from, you know, three or four core systems and a card system. I was like locked up in my office for days. 'cause I just could not. I was, my mind was blown, like this was like early 2000s.
And I think it's still taking a long time for us to get to that point where we can have all that data in one place and slice and dice and really, really become familiar with it. Now we have artificial intelligence to help us, which is good. And we have the whole discipline of data scientists.
And I think the thing that I'm learning is that anywhere that you're, you know, saving in operational efficiencies or anywhere that you're using automation and you're saving a dollar, go hire that data scientist outta college. Go hire that person who's just ready to dig in and help you build those, those patterns and those algorithms because that's where you're gonna win. And, you know, don't hesitate. Just go and do, um, you know, it's an investment upfront, but I think the long game is you're gonna be left behind.
That's yeah, that's great advice. I know that as we've grown, our data science team has paid dividends and our ability to better understand the data that we're, we're working with and the suggestions and insights we provide out to our partners and end users.
That's great advice. Um, so before this uh, session, we were talking Lisa and you had said, you know, data, I always want more data. Um, and so I was like, I was curious. I think that everybody feels this way, right? Of like, man, if we could just have more data, whether it's, you know, pulling from a different system or enhancing the data to find certain insights, I'm curious, are there any data elements that stand out of like, you know, this is an area that the industry still needs to kind of grow into, or that we still see, uh, small amounts of coverage of that data in that come to the top of your mind?
I mean, it consumes most of my days, right? Trying to find, you know, we're so reactive, how do we get more predictive? You know, it's always, you feel like you're, you're learning a day behind and you can go down rabbit holes on what you have access to, but we really need to spend more time. And I think, you know, more collaboration.
When I was in the credit union world, that was the beauty of the credit union. You worked together with other credit unions, you came up with consortiums and opportunities to share and learn and pick up the phone and call someone and say, Hey, are you seeing this? And someone's gonna give you an answer? Yes, we're seeing it too.
Um, in banking, that was a little bit of a shock for me because I couldn't just pick up and call another bank. But I think that that's the opportunity where we have to really become stronger because it is us, us versus them. Yeah. Um, the fraudsters are always, they're always thinking, they're always ahead. And it just is a daily task to make sure that we're pulling in the right points and learning those trends and anticipating it's that whole predictability that we need to kind of get better at how can we be ahead of it for change?
Great answer. Yeah. I think I hear fraud in more and more conversations the more and more time that goes by. Uh, Brian, did you have something you wanted to add as well?
Yeah, so I mean, our perspective is A little bit different. I'm not Saying that fraud isn't, isn't top of mind with, uh, with what we do. But when you look at data across the broad scope of financial institutions, I think you really have to start kind of at a level below the data enhancement is actually the connectivity to the data.
And making sure that you're improving that tokenized connections broadly available across all financial institutions regardless of, of aggregator affiliation or whomever they're partnered with. Um, accessibility to the data has to be key. And then depth of the data is true as well. And this is an interesting thing where you're talking about data scientists and looking at that. So when you're going through the process of adapting and changing a paper-based process into a digital process, you have a lot of things that kind of get drug along from a paper-based process.
So if you think about what's called bank statement lending, and it's a way that it actually exists, people would deliver 24 months worth of bank statements to a lender, wouldn't have to verify income, wouldn't necessarily have to verify your tax information, but they could loan on 24 months worth of physical bank statements. Well now that's gone to, well I want 24 months worth of bank transaction data.
But I think the other side is that, and this is what data scientists will tell you is that what I was doing 18 months ago in my bank account does not represent who I am today. So there's an evolution in the data. So we're always gonna, everybody's gonna always want more data.
But I think the important correlation is to say, what's the right amount of data that you have to have to answer the question that you need? I think investments and how people do their investments is probably different than their capacity for a loan or their capacity to do other things. But you need to tailor the amount of data that you're pulling to the use case that you have. So I think that's an important change that we're gonna have to see as well.
Yeah, fantastic. And data minimization and a lot of conversations that I'm in, uh, seems to always be a part of that conversation. And that's, I think that's how I would summarize the data that you need for the task that you're doing. That's great.
Um, so, we've got about 10 minutes left and I'm gonna merge kind of two questions together and I, I'd like, you know, choose, you can answer both, but, one or the other. Um, so I'm curious, in your roles, whether current or throughout your time working with data in the industry, um, where have you seen data used really well? Um, and then also maybe where have you seen issues with the use of data and maybe cautionary tales?
Um, 'cause I, I definitely think maybe even the data minimization world, there are areas where it's like, whoa, you don't need that data. Why would you let that out there? Right. and so I'm curious, maybe Abbas, let's start with you and work back again. Um, good use of data versus cautionary tales.
Yeah. So, uh, when I think about good use of data, there's two things that come to mind. One is rewards, and the other is suggestions. Starting with rewards, I'm not talking about just credit card rewards, I'm talking about all types of rewards. Your McDonald's app, your REI membership, your credit cards that you've booked to come to MX Summit, rewards has completely changed the way customers interact with, particular transactions.
Um, for example, on my way here,I had to fill up gas. I didn't use a Chase Sapphire reserve. I used my, I used my Chase Freedom 'cause I was gonna get at 5% back, right? Um, if rewards is used in the right way, I mean it, it keeps a customer like locked, right? When you look at REI I paid $20 up front, but I get dividends at the end of the year on all my purchases on 10% back on all my purchases. Um, so again, like I think that's it. It's amazing.
McDonald's, I am one of those guys that love french fries. McDonald's is actually unique. So I'm one of those people that does research on their own and I sign up for everything. Um, so when McDonald's came out for the app, one of my friends, he worked on it. Um, when he was a consultant, he told me about it.
McDonald's tracks the transactions that you do and they target the rewards based on the transactions that you do at McDonald's. The more you purchase at McDonald's, the more rewards you get targeted towards your profile. And I'm embarrassed to say I get free french fries every Friday. So that has become, uh, a date night for me and my wife on Friday nights, and I'm always even a little happy about it, so I can't complain about it. Um, going to the latter, um, suggestions.
Um, when it comes down to suggestions, um, it's, it's absolutely amazing what banking and fintechs can push to you based on your history, uh, of how you surf the web, how you interact with the web. Um, it can even be on like social media sites when you use Spotify and you click on random music, how Spotify changes the algorithm and shoots new music to you that you absolutely love. Right?
To give you an example, I recently just came back from Switzerland and before when I booked the trip initially with United, I'm pretty sure United pushed my purchase and Betterment picked it up and Betterment started marketing their debit card towards me. And one of the rewards, one of the benefits of that debit card was there are no international fees, uh, when you withdraw money from any of those ATMs. And I was a sucker and signed up for it 'cause it didn't hurt me at all, but I reaped in the benefits.
Um, so again, like suggestions can be used for good, it can be used for bad, but, uh, one other thing is, like I mentioned earlier how when I scroll social media, sometimes I come across cat reels and the next five reels are cats and I don't complain about that. But when I look for financial advice and I come across those, financial influencers, I have learned a lot through financial influencers to the point where I would ask my tax accountant about it and I would ask my financial advisor about it and they would gimme a look like, how'd you find out? And it's all social media. Um, so that's where I see data being used for good.
And where I see data used being used for bad is, one thing that everyone needs to take into account is if you ever do any type of data analytic work and there is a sample that takes place when you do your data analytics, you need to be able to define out that sample to make sure there's no bias data.
You know, you come across a lot of articles across the internet, CNBC or MSNBC that come out with these articles that say, you know, average salaries of this and they say a population sample of a thousand oh one, a thousand is not big enough. Two, who makes up that population? Right? Is it everyone from New Jersey? Is everyone from Utah? You have to be able to define out the population sample when you publish something or you're doing data analytics on a specific product that you're working on. And, uh, that's a personal experience.
We, I was building out payments product in my past and one of the questions I asked was your, the model that we were using at the time, um, what's the population sample? 'cause there's a lot of outliers in these right now. And they did not think about that. They did not think about the biases in the data and we had to refactor the whole entire model and start from scratch. And so that's where I see the good and the bad.
Yeah. Going off of that, the biases that are either in the data or they're introduced by the people analyzed in the data is very important. And the lending space, there's a con there's a concept of adverse impact, which means that I might be denied a loan for some reason that isn't necessarily a person with the exact same financial capabilities would be approved for that loan.
So you have to understand not only what makes up the data, but what is being, what is going into analyzing, telling the story about that data so that you don't introduce these adverse impacts. And then you get into the whole process of adverse outcomes, which is a whole nother story that where you can take a 100% accurate view of data, you may still have adverse outcomes from that. So that's a challenge that a lot of lenders have had with the use of direct source data.
Um, here's another one, and it kind of goes off of the data minimization, space is lenders only want to see as little data as they can. If something comes into their house, they have to analyze it. So let's say they got three months worth of bank statement history and on day 75 there was a $50,000 deposit in their account. Well, their under underwriting guidelines say you only have to look at the last 60 days, but because that report had 70 had day 75, a $50,000 deposit, now they have to go to the borrower. Where'd this money come from? What was the source of it? Did you have a gift letter for it? I need to see the account where that came from.
So it introduces challenges operationally for the lender to be able to do that. So one of the things that we've done with account check is actually deliver multiple views of data. So the automated underwriting systems that will look at the data, digital data digitally up at Fannie Mae and Freddie Mac, they'll consume as much data as we have and they know where the gateways of data reside. They'll analyze the first 60 days using these set of guidelines and they'll look at the rest of the data with other set of guidelines to expand the opportunities for this loan to qualify the lender. They're looking at the last 60 days 'cause that's all they want to see. They don't wanna see any more data for that. So looking at the other side, the areas for good and I talked about the rental payment program, and that's been a huge success, uh, for expanding home ownerships.
But you get down to the more personal, and this was back in, in the earlier days of, of account check when As a CTO is also taking help desk calls. And so you've got that, that loan officer with the borrower right across the table and they need to refresh the data in the account because they gotta see that payroll deposit that came in yesterday so they can close on the loan today.
You walk them through the process, Hey, pull up the portal, select the order, click refresh. 30 seconds later they have the report. It shows the, it shows the direct deposit, they qualify for the loan and they get their home. Those are the success stories that really make the story around direct source data and access to data that much more tremendous in how it can affect people's lives.
Very cool. Yeah, this is awesome. I think, for all of us who've gone in and gotten home loans, that sounds like a fantastic experience. Uh, Lisa, do you have anything to add from a lessons learned or data used?
Well, perspective, I think the more you can use personalized data, the more you build trust, the more you build trust, the more they feel secure, the more they're going to come back. Um, story after story where, you know, someone has gone through the effort of setting the goal of home ownership and then seeing that come to life through, you know, the ability of either digitally presenting offers that put them in the right space, pri providing products to put them in the right space, watching them grow and mature and end up with something as grand as home ownership or, you know, starting a new business. You know, those stories are over and over and over again. And the more that we can use that data to put those stories together for our employees and our customers, I think that that's the reward. I don't wanna talk about the negatives though.
Okay, no worries.
But I think the rewards though, I meanI think that has been kind of like the sleeper, you know, I don't think we learn enough about using the rewards game and especially like, I have a daughter who's starting in her career and you know, she's building and she doesn't see home ownership as something like in her purview, right? But what I tell her is like, you can use rewards to augment your lifestyle and if you just look and play, play the game and educate yourself and there's, you know, lots on social media help you get there.
I think that's how we give them hope that, you know we can free up some of that income that they can start saving for their home ownership. So yeah, And rewards will change behavior and open up opportunities, just like you're saying, it's, it, it is a game changer.
Um, thank you for the conversation. I know that we're at time a little bit over. Um, thank you. It's been a great conversation. I look forward to all the conversations that come out of this, throughout the conference. Um, can we give the panel a round of applause and, uh, thanks.
I do wanna leave off with one more thing. Nate, what's The Work on a data enhancement product? Remember in the end, data is just data. You need to create an actionable step for your customers. That's the most important thing.
Thank you. That's a great point to end on. And, um, what I will say, if you are interested in anything that's being done, by the different organizations up here, please come talk with us. Always glad to talk about data. I think, uh, for those on the panel, it's, the lives of what we do and so, um, yeah look forward to any questions that you have. Thank you again for the time.
View Full Transcript