Embrace change, take risks, and disrupt yourself
Hosted by top 5 banking and fintech influencer, Jim Marous, Banking Transformed highlights the challenges facing the banking industry. Featuring some of the top minds in business, this podcast explores how financial institutions can prepare for the future of banking.
How Innovation Can Drive Growth in Lending
As the lending landscape continues to rapidly evolve, financial institutions must embrace innovation and collaboration to remain competitive. In this special mini-series on the future of lending, sponsored by LoanPro, we are joined by Charley Ma Co-Founder & Managing Partner at Exponent Founders Capital and Colton Pond, Chief Marketing Officer at LoanPro. They discuss the importance of innovation in lending, emphasizing the difference between noise and emerging opportunities.
Jonas and Colton also share their perspective on the potential impact of artificial intelligence in lending and highlight examples of companies that are successfully leveraging AI to revolutionize financial services.
This episode of Banking Transformed Solutions is sponsored by LoanPro
LoanPro is the leading modern lending and credit platform enabling lenders to innovate quicker, driving growth while optimizing efficiency. Over 600 lenders use LoanPro to enhance their borrower, agent, and back-office experiences. LoanPro’s mission of providing the platform to innovate the future of finance is enabled through its composable architecture, allowing lenders to enhance their origination, servicing, collections, and payments, all supported by a modern lending core.
Where to Listen
Find us in your favorite podcast app.
![](https://polish.evergreenpodcasts.com/shows/_650x397_crop_center-center_100_none/Jim-Marous_Gray-coat-arms-folded-disrupt.png)
Jim Marous (00:07):
Welcome to another episode of Banking Transformed, the podcast that dives deep into the trends, innovations, and strategies shaping the future of banking. I'm your host, Jim Marous, owner and CEO of the Digital Banking Report, and co-publisher of The Financial Brand.
Jim Marous (00:23):
As the lending landscape continues to evolve, financial institutions must embrace innovation and collaboration to remain competitive. In this special mini-series on the future of lending, sponsored by LoanPro, we are joined by Charley Ma, co-founder and managing partner at Exponent Founders Capital, and Colton Pond, Chief Marketing Officer at LoanPro. They will discuss the key trends and themes in credit and lending, emphasizing the importance of distinguishing between noise and genuine opportunities.
Jim Marous (00:58):
Charley and Colton will share their perspectives on the impact of artificial intelligence in lending, and highlight examples where companies have successfully integrated AI to revolutionize financial services.
Jim Marous (01:10):
As technology advances and customer preferences shift, banks and credit unions must embrace new ideas, technologies, and partnerships to transform your lending processes and offerings. One of the biggest impacts will probably be the result of AI integration driving significant innovation opportunities and transforming the products and experiences for borrowers.
Jim Marous (01:33):
So, Charley, let's start with you today. As an investor, what are some of the most promising collaboration opportunities you see between traditional and non-traditional providers in the lending space?
Charley Ma (01:45):
Yeah, it's a great question, but first of all excited to be on the podcast with you all, it's an honor. Long time reader and fan, Jim, of everything you've put out there for the industry. So, excited to chat and Colton, good to see you as always.
Colton Pond (01:58):
Thank you.
Charley Ma (02:00):
It's a good question. I think right now we're entering an interesting time as it relates to kind of fintech partnerships and banking partnerships. One of the things that I've been talking about or at least what I've sort of seen across I think the B2B fintech landscape is that, it felt just to me the last five, six years or so of "investment in fintech" was really kind of fintech infrastructure built for fintechs or built kind of for the disruptors.
Charley Ma (02:31):
There's a lot of these neobanks, there's a lot of new potential kind of early stage, fintech companies geared towards businesses, towards consumers, and a lot of the investment and infrastructure opportunities was being invested in companies that were enabling those type of companies. Like kind of the, "Just represent the gate."
Charley Ma (02:48):
And it makes sense, I think that if you're a larger financial institution, typically more risk averse, which I think makes a lot of sense and want to see these use cases really proven out in the field. But what we see now is that I think as the fintech industry has matured to where we are now, we have started to see fintech companies get to real meaningful scale that is quite impactful.
Charley Ma (03:14):
And as a result, the infrastructure that a lot of these companies are built on top of are quite robust. So, I actually think we're now entering a bit of a golden era I think around sort of fintech partnerships between banks and fintech infrastructure providers. But we're also seeing, I think, a lot of regulatory headwinds.
Charley Ma (03:33):
Obviously, I think there's been a lot of news on ... this whole thing is around sort of banking as a service and what's happening in the ecosystem around sponsor banks there, which I'm sure we'll get into at some point.
Charley Ma (03:44):
And so, I think it's both an opportunity for best in class, infrastructure providers across the fintech landscape to start to actually be able to partner with financial institutions that have been around for much longer. And it's one of the areas that I think I actually have been quite interested and excited about, it's taking this a few years, is particularly around, I would say, lending and underwriting.
Charley Ma (04:09):
I think we're starting to see the adoption and usage of real time data incorporated into various decisions. Probably, I think a lot of the advancements initially started off in SMB lending and we're now starting to see that, I think start to come to effect in consumer lending, but happy to dive in further.
Jim Marous (04:25):
So, sticking with you for a second, Charley, there's been so much activity in the payment space, obviously. Do you see that activity then becoming a lot more like a lending product because they're really not that distinct in many ways, even though they're very different in the way you implement them, but Buy Now, Pay Later is a payment product that really is a lending product as well. Do you see a lot of that really making it so that the lending space is going to get hyperactive?
Charley Ma (04:53):
It's a good question. I think what we're starting to see are more and more companies wanting to get access to actual consumer and businesses, and it turns out — I've been in fintech for a very, very long time. I love working with them and using fintech products, but I still bank with a good old-fashioned bank.
Charley Ma (05:18):
I think there's still a high degree of trust that you give towards a banking institution. I think there's been a lot of marketing and say on, you want an all-digital experience, and you only need a mobile app. And I think at the end of the day, when it comes to complex decisions, you still want to talk to someone that understands your background. Someone that's able to have a deep sophistication around the financial products they have.
Charley Ma (05:40):
I still want to go to a branch. I still want to call in to somewhere and talk to someone about a complicated financial decision. And I want to talk to someone that has a good understanding of the relationship that I've had with that institution for a long time.
Charley Ma (05:54):
And so, I think that a lot of the movement across, for example, with Buy Now, Pay Later, I think we're just starting to see early signs of this. There's just a lot more partnerships with networks and ecosystem that have access to a large base of businesses, a large base of customers. And that could be a bank of financial institution. It could be a large merchant; it could be other providers.
Charley Ma (06:15):
But I think that sort of partnership model and distribution and I guess other people, you can call it embedded fintech, but I view it as just that, it's just another form of distribution that a lot of people are kind of starting to partner up with.
Jim Marous (06:30):
Great. And Colton, from your perspective, I've said often that in order to innovate, you need the back office to be fixed. You can't innovate a great product and have a legacy operating system because eventually it's going to break. It's just not going to be as efficient, not as seamless, not as overbearing from a consumer standpoint. How does LoanPro help with the products you have to really provide the platform where innovation’s possible?
Colton Pond (06:58):
Yeah, great question. To your point, I think to disrupt the next wave of innovation. I love Charley's point on banks and credit unions have a significant amount of trust. And I think we've even seen that with the larger institutions. If you looked at Silicon Valley Bank and the failure of Silicon Valley Bank and where funds flowed, J.P. Morgan got a big bulk of funds and then Brex got a big bulk of funds, but those funds were just like, take quickly, go to Brex because they had quick, instant payments, and then go somewhere else.
Colton Pond (07:30):
And we see lots of big trust in these financial institutions, and there's a huge opportunity for banks and credit unions to garner that trust and to create modern, unique experiences on top of a modern infrastructure, and to slowly modernize their infrastructure on top of their core banking platform.
Jim Marous (07:49):
So, when you're looking at organizations today, you're seeing a really strong trend away from build to either a buyer collaborative experience where they're either buying products from a, let's say a fintech provider, or even more importantly, collaborating with those organizations such as LoanPro that actually gives them the foundation, the ability to innovate going forward.
Colton Pond (08:12):
100%. And this goes back to what Charley mentioned on previously, fintech was very focused on enabling other fintech. And we've seen, especially over the past year and year and a half fintech focused on how do I enable banks and credit unions?
Colton Pond (08:26):
And to your earlier point, it would be remiss if I don't mention it on payments. I do see the world in which payments in lending and credit become more similar and more combined. Visa announced their Visa payments form their flexible credentials product, which is not available in the US yet, but they said that there will be pilots late 2024.
Colton Pond (08:49):
But that essentially is, I have a card, this card is an access point to any access of funding I have. That may be my debit account. It may be checking, maybe savings, maybe consolidating my wealth account, my 401(k), selling something from my wealth account. And to be able to pay, it may be a line of credit, it may be a credit card, it may be an installment loan, it may be a home equity line of credit. And in real time, I'm using a card as an access point for all these payment methods.
Colton Pond (09:15):
And that's the future of combining lending and credit together or phone, app, is probably a better way to put it today. But the idea is, as you modernize your infrastructure, those are the types of things that are possible. Where in today's infrastructure of these legacy infrastructure that's 30, 40, 50, 60 years old, those experiences aren't possible, because my core system or infrastructure doesn't support it.
Jim Marous (09:43):
So, to that point it's over a decade ago. I had a friend who worked for a Polish bank, and they build a brand new mobile platform that allowed almost everybody, I think it was 80% of their customer base, to get some version of a line of credit and had it made it so you could access it either through an overdraft, if that was the case, or by an additional push of a button on their mobile banking app.
Jim Marous (10:07):
Now, that took some innovative thinking. Number one, you were looking at credit in a different way, because it wasn't about a 1000, $2,000 line of credit. It was everything from a $100 line of credit up to those kind of levels.
Jim Marous (10:19):
In addition, instant access. So, you are continually evaluated based on your most recent credit bureau that’s instantaneous. But the flow of data, the flow of between organizations, between product lines, was such that if you wanted to move credit into a savings account, if the rates were inverted or something like that, you could do it.
Jim Marous (10:38):
What amazes me is that, very few, if any organizations today offer that all in one credit availability to virtually their entire base. And part of that has to do with senior management within the company, never embracing the fact that a small line of credit can make them money if you do it across the whole base, because you're going to retain those customers.
Jim Marous (10:59):
How does that type of innovation happen? And how does your platform actually support, as you just mentioned, the integration of all these products into one that's easily accessible and basically seamless between products?
Colton Pond (11:14):
So, to your last point, LoanPro is the modern, I say the modern lending credit platform to launch any class of loan line of credit and credit card. So, there's a really cool synergy between, I have a credit card and I swipe my credit card and it's $5,000 at Home Depot. I'm probably doing home renovation.
Colton Pond (11:32):
And if I'm a bank, I should probably ping that customer and say, "Hey, would you like a HELOC? And would you like to roll that $5,000 transaction? You just made it at Home Depot into a home equity line of credit?” And now I have my home equity line of credit there.
Colton Pond (11:44):
To your earlier point though, on expanding the availability of credit, we actually have a customer, one of my favorite customers, Aero Finance, a small fintech that does credit cards. And they're famous because they offer the smallest line of credit available today, and it's a $50 line of credit available to thin file no file folks.
Colton Pond (12:04):
And what they found a way to do is they offer that $50 line of credit attached to a credit card, and then if you make it on-time payment, it goes to a hundred dollars. If you make three on-time payments, it goes up to $300, and you do certain actions and they infer your actions, then they increase or decrease that available balance in that line of credit. And eventually the hypothesis is, that's going to grow, and they're going to become very profitable customers.
Jim Marous (12:27):
So, you're going to have an overdraft in your checking account, you pay it off with a credit card, you're not going to be penalized at $35, but you're going to get another $50 line of credit on top of the 50 you already had. It's kind of like a win-win, that you get out of trouble, and you're given something in return.
Jim Marous (12:43)
Charley, from your perspective, there's no lack of great ideas out there. Certainly, in the financial services space, there's a lot of great companies out there. What gets in the way of financial institutions saying yes or actually implementing them, the other apps that are buying it and then implementing in a clunky way that isn't really the way it was meant to be presented?
Charley Ma (13:06):
Yeah, that's a good question. Something that I think a lot about when we're talking to founders that are building and trying to partner with financial institutions, is that got to realize that interoperability is extremely critical into the stacks that they're using and the current workflows and processes that currently exist within the bank.
Charley Ma (13:26):
When you're working with a bank, it's not, hey, integrate this API and we're going to be good to go tomorrow. It's, well, no, you have to be able to really understand what did the branch bankers do? What are the underwritings do? What does the back office do? What products and tools are they using? How do you actually embed and make their lives better? So, actually, it requires a lot of investment and time into a partnership.
Charley Ma (13:50):
And that's where I usually see things go wrong. Where there's failure modes on maybe the bank, CEO or Chief Innovation Offer comes in and brings in a, hey, this is a, this, we got to use this brand new AI tool, I want to deploy it. I signed a big contract, and we're pushing this out. That’s a really, really big failure mode.
Charley Ma (14:11):
There's another one on the fintech company or the infrastructure provider comes in and they're really excited and, hey, we want to go live in a month. And they haven't really invested the time and effort to really understand what are all the different processes. What are the key integrations?
Charley Ma (14:24)
How do they make sure that they actually have interoperability within all the different blanks platforms. To actually drive successful implementation and actually drive results for what actually matters, which is the end customer and what you're actually doing for the end customer member of that financial institution.
Charley Ma (14:40):
And so, usually it's often more so its funny, founders complain all the time of ... well, it takes forever selling to a bank. But there's a very good reason why we're, at the end of the day too, banks are regulated, and you can't really make a mistake, if you make a mistake, it's different than when you're selling I don't know, selling some productivity tool. If you make a mistake, no one's losing money off of that.
Charley Ma (15:07):
But when you do make a mistake with something that is touching someone's money that is a lot more drastic. And so, as a result, there's just a lot more effort I think that's required to activate those partnerships. But on the flip side, if you're able to build those trusted relationships and invest the time and effort, like those were, where some of the best customers and partnerships really start to emerge.
Colton Pond (15:26):
Charley, I love your take and feedback. What I love about your role throughout your career is you essentially spent time at three of the most iconic fintech companies in the early days with Plaid, Ramp and Alloy, and building out the go to market, what that looks like. And then today, as an investor, you see a lot of fintech companies.
Colton Pond (15:48):
I'd love your take on advice you give to fintechs on how to ensure partnerships with banks and credit unions are actually successful. I read a stat the other day that 70% of partnerships that are actually signed with a bank and fintech never go live. What's the insights that you've seen across fintechs that you've worked with that ensure those go successfully and help enable innovation?
Charley Ma (16:16):
Yeah, it's a great question. I think something that we spend a lot of time talking about is, building a product for a financial decision for a bank isn't like building a "Regular startup." You can't move fast and break things, just roll this out and experiment and we'll see what happens.
Charley Ma (16:38):
You need to have a product that is extremely robust to mistakes. Which is usually actually a function of time and just a lot of product or robustness being built and being able to be patient with those processes. And so, I think that oftentimes there's ... I talked to a lot of founders on just having the right mentality around what does it mean to actually work and partner with a bank.
Charley Ma (17:00):
I think the vast majority of banks, they're not viewing this as this is an option in the next two years. We're going to look at another provider and swap it out. Generally speaking, when you select a provider you're selecting them for multiple years, if not multiple decades.
Charley Ma (17:15):
And so, as a result, it's almost like a marriage, where you're going to want to spend a lot of time with that vendor. You want to have a lot of case studies and proof points. You want to understand what's the impact you're going to drive.
Charley Ma (17:28):
And you want to have a high degree of trust that10, 20, 30 years from now. If you're a core part or critical part of my infrastructure I could depend on you, you'll continue to be around. Which is why I often think that the hardest thing as an early-stage company is solving that chicken and egg problem set.
Charley Ma (17:46):
And so, I often talk about, well, how do you start off with, at least on the initial use cases maybe more lightweight use cases that are more robust to mistakes. If you're coming out of the gates and saying, I want to replace your banking core, it's not going to happen.
Charley Ma (18:02):
The number of founders that I've met that says, well, I want to go after the FIS, the Jack Henrys, and I'm going to rip out all these on-prem cores, we got to move all these banks to an On-Cloud platform, and I always go, good luck with that, it's not going to happen. That is the third rail that you don't want to touch because so much is dependent on that grid infrastructure. If something goes wrong, that is potentially business ending.
Charley Ma (18:30):
And so, early on, we talk a lot with our founders about how do you start off with wedge use cases that can quickly demonstrate a lot of value to that bank while also being robust to mistakes. Some of those tend to be, for example, in customer support and on kind of handling tickets there.
Charley Ma (18:52):
It could be more of a, I think what we're seeing, for example, with AI, a lot of kind of co-piloting type use cases on how do you help operational teams, back-office teams to be more efficient. In the case of LoanPro, it's how do you actually handle document ingestion? How do you actually handle digitizing all the processes that should be, and just making the bank more efficient and operate better for the end customer. And those use cases tend to be, I think, really, really good ones to just start off with.
Jim Marous (19:17):
So, Charley, what have you seen at your company that excites you in the whole realm of lending innovation and besides LoanPro? What organization has really picked your interest and said, it's a pretty good solution.
Charley Ma (19:39):
It's a great question. I mean, I'm always happy to shout out the previous companies I've worked at, those are the ones that I know the most. I think for example, in Plaid's use case, I think we're now starting to see the adoption of open banking data in the United States and on kind of consumer permissioned transaction data incorporated as part of underwriting decisioning. I think we're still in the early innings of that use case.
Jim Marous (20:04):
But I'll take Ramp for example. I think that was one of the key innovations for us as part of underwriting, was being able to connect into a wide variety of different data streams about a business to get a more comprehensive understanding when we're underwriting it.
Charley Ma (20:19):
We're not just relying upon ... historically it'd be, if I'm underwriting SMB, I would pull in the credit score of the beneficial owner and maybe a little bit information about what the business does and that's pretty much it. Now, there's a lot more data out there, there's a lot more ways to get access to, more and more data about that business. You can have a much more comprehensive understanding of that business and underwriting that.
Charley Ma (20:42):
And then finally, with Alloy, my previous company, for example, one of the products that I helped build there was our onboarding decision engine and being able to incorporate a wide variety of different data signals, both at onboarding and also at sort of ongoing transaction monitoring to be able to catch fraud at the onset and then leveraging AI machine learning to pull in a variety of different data signals and detect, one, would they be potentially sort of breakout fraud across your customer base.
Charley Ma (21:07):
And then the key thing for all of those was interoperability into the core infrastructure providers that actually matter for those financial institutions.
Jim Marous (21:14):
You know what Charley, great example, I'll use Alloy because I have a question for you. Alloy, we work with a lot of companies that want to find that perfect new account opening experience, that front end. And I've had Alloy people come up to me and go, my example is, spot on in that the biggest problem with Alloy, is bankers that mess it up where they say, yes, I want to use this, this, this, but I want to start with a driver's license. And it just breaks what was built.
Jim Marous (21:43):
From your perspective. Number one, what have you seen in the marketplace? Because you have a lot of clients, but you're still a new fintech. What have you seen in the marketplace that you say, this is really going to determine whether or not it's going to work or not, where the relationship's going to work.
Jim Marous (21:57)
So, what is a key element that says, I feel pretty good about the success of my implementation because of this? And then the alternative, what is the one thing that kind of gets say, oh, geez I sold it, but in my back of my mind, I'm not too sure if it's going to go live. What are those elements?
Colton Pond (22:14):
So, I think there's two to call out. One is and I do some angel investing, not nearly as much as Charley. I want to get to the point that Charley is. Charley's like the state that I want to get to. But in talking with a lot of fintech founders, there's a lot that have interesting ideas. Most have interesting ideas. And most think, yeah, I'm going to sell to banks and credit unions.
Colton Pond (22:38):
One of my first questions is, do you understand how banks and credit unions operate? Do you understand their needs? Do you understand what problems they're going after? And it goes to a conversation we just had on understanding the customer. If you're a fintech and your customer needs to be banks and credit unions, how do you understand them? How do you make sure your product actually solves a real need that they have?
Colton Pond (23:02):
And today, in the current banking environment where deposits are shrinking, interest rates are high, and it's causing banks to compress and try and improve their margins. They are only spending money on things that actually fix problems and actually drive value.
Colton Pond (23:18):
The second aspect is, knowing the underpinnings of a bank and how many people you need to get consensus with. I've seen lots of partnerships where they get consensus within one group and they sign it, maybe it was the product group, but the IT group had no idea about it, and the marketing team had to do something, had no idea about it, and then after signing it, they got to take a whole step back and be like, now you got to go through this whole kind of like sales process of IT and marketing and maybe they have other ideas.
Colton Pond (23:50)
And it's the importance of driving breadth of cohesion and awareness around the impact you're trying to drive at that financial institution.
Jim Marous (23:58):
How about the financial institution itself? What aspects of the financial institution are key for the ability to turn on something new? Because we know change sucks, I say that I'm thinking every podcast now, but the bottom line is no one's dying to change what they currently have. Nobody's trying to get out of Jack Henry or FIS or Fiserv, but they're still looking for new things.
Jim Marous (24:21):
From a financial institution perspective. As you're meeting with them, what is the key element they've got to have in place in order to make that new solution work? And then the alternative to that, very similar to what the last question was, is what is the one thing that you go this finance institution, it's not going to go because of this. What's the positive, what's the negative from the financial institution side?
Colton Pond (24:43):
My recommendation generally is how do you do crawl, walk, run? So, oftentimes banks and credit unions can overcomplicate it. Well, it's got to integrate with this system, and it's got to integrate with Fiserv, and it's got to integrate with this and this and this. And it's like okay, maybe, but how do you look at things outside of that? How do you launch something on its own to test and prove that it works, and then expand the use case?
Colton Pond (25:07)
That's the most successful banks and credit unions I've seen. Lots of them have started to stand up a de novo digital bank, where is there innovation hub? It's like, hey, we can test new products here and we can see if they work, and then if they work, we can roll them out to the mothership and drive the value from them.
Jim Marous (25:24):
And by the way, if they don't, we're going to fill — because companies will say, yes, we want to do this, and if it works, we'll roll it out. But are you willing to turn it off? And that's a hard one because that's not traditional banking, but traditional banking is, I'm going to test the living daylights out of it. I'm only going to make something go live if I know it's going to work. Sometimes it's better to not have that happen.
Colton Pond (25:45):
One of the problems though, and the reason being is if you rely on your core and you think everything needs to be connected and integrated, say if I serve my score with Fiserv, it will take you,12, 18, maybe 24 months to launch that.
Jim Marous (26:00):
If you’re lucky.
Colton Pond (26:01):
There's a lot of time that you invested into doing so, just to see if something works. But if you are able to find innovation ecosystem where you can test out new ideas and products in a safe way and quick, fast, and easy, then you can actually see value, no value for us, and let me roll it out and drive more value.
Jim Marous (26:23):
So, Charley, not many conversations go on nowadays. Somebody bring up generative AI. How do you see generative AI changing the lending process or the ability to lend, which I'm looking that was a little bit different, not the hard stuff, but the soft stuff as well.
Charley Ma (26:42):
Yeah, I think we're still definitely in early innings on LLMs, and I think that, we’re starting to get a better understanding of what are LLMs actually good at. And also, I think the downsides of, there's a lot of failure modes with LLMs.
Charley Ma (26:57):
I think generally, I tend to bucket what are LLMs actually good at. They're actually good at text generation based off of given specific prompts. They're good at processing vast amounts of information and kind of summarizing they're pretty good at information retrieval across vast amounts of information.
Charley Ma (27:14):
And they're very good at things like co-generation and turns out they're always off. But there's a lot of things that we now realize that LLMs are not good at, they make stuff up, they hallucinate. Depending upon the training data, there is bias and fairness issues.
Charley Ma (27:32):
The output is non-deterministic. If I told any of those things, those things are very bad when it comes to lending. And so, as a result, I think that we're still in the sort of early innings on really leveraging LLMs and production use cases as it relates to underwriting but we're starting to see at least I've been starting to see some interesting, exciting use cases.
Charley Ma (27:54):
So, first of all, I do think that a world where we've reached AGI and it’s replaced us. I think we're a long ways away from that, at least with definitely within financial services, where a lot of the more successful deployments of LLMs within FIs, I think have been as more of a co-pilot type structure. Where you're augmenting an existing human capability.
Charley Ma (28:17):
Whether it could be your underwriting team. And it's, hey, it turns out when you're underwriting an SMB, you have to process a hundred documents. We already know that you have strict SLAs when you're underwriting a loan because you want to have a really seamless customer experience.
Charley Ma (28:33):
So, as a result, there's no way that you're going to process a thousand pages of data with a person reading through it. You might process the most important documents, and that should always be done by a human, but there could be a lot of other extra information that you can incorporate in that loan.
Charley Ma (28:47):
Some of it may be extra, but some that actually could be really, really critical. Great, that's actually something that an LLM could be really fantastic on. Identifying extra signals that could be high signal for you as part of an underwriting origination system.
Charley Ma (29:00):
Document processing, it tends to be quite more, I think, a lot more deterministic, and you could put a lot more boundaries and controls around it, and it's a lot easier to curate. We've seen a lot more use cases around there where LLMs versus, say, a more traditional OCR technology has been much more successful in deployment there.
Charley Ma (29:15):
And then I do think early signs around kind of customer support and servicing, where if you have to do dialing with an offshore team, if you have to do collections with an offshore team a lot of those could be done with a very well-constructed LLM that's specifically built for that use case. But we're just starting to see some use cases that emerge around there.
Charley Ma (29:40):
But I still land at, there needs to be a lot of supervision required because of how nondeterministic a lot of these LLMs are right now today.
Jim Marous (29:49):
So, Charley, I'm going to take a question I gave to Colton and give it to you as well. We sometimes think that the field of dreams, if you build that, they will come. And there's a lot of great solutions out there that have fallen on their own way because people didn't know how to implement them, or organizations didn't know how to sell them. There's a lot of other instances.
Jim Marous (30:08):
You've been at your previous organizations all have done an extraordinarily good job of deploying, at least looking from the outside, looking in, deploying in a way that was meant to be deployed. In other words, they brought value, they got bought, they got implemented correctly to some degree, at least, and they succeeded.
Jim Marous (30:29)
But there's a lot that had been left on the cutting room floor for one reason or another. What gets in the way between innovating and building and deployment that really, we sometimes miss in the whole innovation process.
Charley Ma (30:47):
I think one of the kind of classic traps that I see is founders or executives that sort of fall in love with the vision and sort of lose sight on what actually works or what actually will get implemented within an institution. Where we can talk all we want about I think there's a lot of boardroom discussion on AI and how it's going to replace humans on manual operational tasks and all that stuff and that's great to talk about in theory.
Charley Ma (31:16):
And I think if you spend too much time talking at more maybe that exact board level and you don't actually spend the time talking to your team members, talking to actual customers, talking to whoever's actually doing the work and understanding, okay, how would it actually drive efficiencies? How would it potentially replace what the work is doing? What end processes that actually affect?
Charley Ma (31:36):
If you don't actually take into account the actual end workflows that you're trying to inflect, trying to improve or trying to completely change, I think that tends to result in products that don't actually make a lot of sense. Where on paper sounds great, the demo part looks amazing, and then when it comes to actually getting value from that product it falls very, very short.
Jim Marous (31:58):
Colton great, by the way, Charley, great answer. And we take a lot of that for granted, and that's where the gap is. You look at Plaid, it's relatively involved product. And as is Alloy and ... it's interesting because we can, in our minds, oversimplify and miss the biggest possible gap. Speaking of gaps, go ahead.
Colton Pond (32:23):
To that point I was going to mention, you even see it Charley in Plaid, for example, there was a long time where Plaid was not a CRA, and it was less focused on lending from some of the things that Zach and others mentioned. But you've seen a significant, I wouldn't say shift, but probably expansion over the past two years.
Colton Pond (32:43):
Now they're CRA, they did this agreement or partnership for cashflow underwriting with present data and others. And your point is spot on. I think one of the biggest failures happens when you get tunnel vision and like, I have the vision, I'm going to go achieve it, relative to let me hear with market feedback and what the market is telling me to do.
Charley Ma (33:05):
A hundred percent.
Jim Marous (33:06):
So, from your perspective, we at the end of the day, have to think about the consumer, the end consumer. You're in the marketplace all the time. You're talking to organizations, you're in the lending area. Yes, you're talking about your product, but you really, I know your habits pretty good. You're asking a question about the entire realm of the lifespan of a loan. What's the biggest gap we have today from the consumer standpoint between what's possible and what's being delivered?
Colton Pond (33:33):
You asked a great question. So, first I'll mention one thing, and then I'll get to your question. You're right, when I joined fintech back in 2019 one of my first jobs is to go travel the U.S. with, at the time our co-founder or the organization I worked for, co-founder, CTO, and go visit with banks, credit unions, fintechs. And I would always ask the executive, what are your priorities? What are you focused on? What does it look like and why?
Colton Pond (33:59):
And back in 2020, 2021, everyone was like SMB banking, SMB banking focused on small business. And Charley mentioned it. But a little hot take on my side. We've seen very little innovation come from banks and credit unions and small business banking over the past three years.
Colton Pond (34:18):
They've been talking about it, but not doing anything. And the more I talk to banks now in credit unions, 2024, I see them starting to do things about it, which really excites me, because previously small businesses have either been lumped in, oh, you're a consumer, let me treat you like a consumer, or you're a commercial and they're neither.
Colton Pond (34:41):
Some of them may be a single member LLC, but it's not their consumer finances. They're trying to run a business. They have different goals. Different jobs to be done. They're definitely not a commercial client like McDonald's or Walmart or one of those.
Colton Pond (34:55):
And to answer your real question is you need to understand the segment you're going after and try to find ways to attract that segment in the right way. Launching products, modernizing your infrastructure.
Colton Pond (35:14)
I personally believe one of the biggest opportunities is, folks have focused on origination a lot over the past four to five years. Almost every banker credit union now has an online origination flow and process, and a decision engine from someone like Alloy, or maybe it's nCino or someone else.
Colton Pond (35:35):
Where there's been much less focus and the reason why I came to LoanPro is on the loan servicing collection side, how do you build a best-in-class experience, touching the customer, touching the member in unique ways?
Colton Pond (35:46):
And I love Charley's point on looking at ways that new technology like AI can play in that customer experience. I've seen some really cool things from folks like Airbnb, where Airbnb now has a virtual assistant where if I'm a host and I have problems, I can video chat with this virtual assistant that has facial expressions and everything.
Colton Pond (36:09):
And when I ask her a question, she's able to retrieve the information and give it to me in a way more efficient way. And I think there's a lot of unlock on how do you service the customer and the member in the right way?
Jim Marous (36:22):
Essentially, I'm going to take that a step further and talk about your origination side saying, yes, we have 80 some percent that now can take an application digitally. The challenge is, but is it fast and is it easy, and seamless? And it's not. So, the gap and I think is important is we got to do the basics exceptionally well.
Jim Marous (36:42):
I look at Apple card and go, the application for the Apple card was a four-step digital process where I only had to give my Social Security number, my annual income. I had to look at my personal debt and say, is it right already all prefilled?
Jim Marous (36:56):
And then how much money do you want? We got to get it down to that simplicity to a DIY perfection where we get the basics perfect. And as you said, fix the back office so we don't step all over ourselves after we've gotten the loan.
Jim Marous (37:12):
Charley, as we kind of wrap up what do you think is the major thing we need to do as financial institutions and as fintech providers to make lending better, going forward? What is that gap and what recommendation do you give to both fintech companies and financial institutions when it comes to innovation?
Charley Ma (37:32):
This is a super simple answer on my side, but I think just continued discussion in forums like this, where I think that particular at the exact level, just being able to talk to and understand what are the capabilities that could be possible that are maybe four or five years out, but just start to think about what does the realm of possibilities start to look like.
Charley Ma (37:54):
And I actually think the other part is highlighting actual successful use cases within banks and FIs. I think that a lot of times if a bank finds a successful partnership, they'll kind of keep it secret. It's almost as I'm trying to keep this alpha contain within me, and I want to bear hug this amazing thing. I don't want to tell anybody.
Charley Ma (38:13):
And I think that it ends up — I actually think that most FIs are actually more cooperative than not, when it comes to their customers and the customer base. And I think just having more of that dialogue and discussion of actually successful limitation of partnerships and use cases, and having more of those on the world just helps us actually understand where actually can we actually innovate and bring new products and solutions.
Colton Pond (38:39):
And how can we learn from other peers on what they've done and how they're doing it. Along those lines, have one other quick question for you, Charley. Because you see so much in the fintech and financial services space, outside of AI, and everyone wants to talk about AI, and you gave, the best answer that I've heard in a long time about AI and how to actually action on AI. What are a few other trends that you're seeing within banking and fintech in general that folks should be paying more attention to, but they're not?
Charley Ma (39:11):
I a hundred percent still agree with Jim. I still think that we kind of got the first wave of digital count onboarding, and everyone has a digital application, and we've made it easier, where you don't have to go to a branch now, for a lot of providers, and you can do it, do it online.
Charley Ma (39:28):
But I think the consumer expectation, whether you're running an SMB or even just a consumer opening account, they want that Apple Pay experience. They're like, hey, apple is offering me a card, and it takes — literally, I can get the card provision and start spending instantly and that is the bar that I think consumers are looking for.
Charley Ma (39:47):
And so, I do think really nailing down the basics, and it turns out Apple's always been right in that sense of nailing down the complex ... making it look really, really simple is really, really complex and a dozen details there.
Charley Ma (40:00):
And then I think the one area that I am actually interested around AI is, I feel as though every three years we talk about this dream of extreme personalization of product experiences, of offers, of relationships from the banker side to the end customer.
Charley Ma (40:18):
And I think that with the admin of LLMs, it'll actually allow for that to actually happen. Where you can actually instead it was — I can't afford to have a banker dedicated to every single person at the bank. And it's only qualified for my ultra-high net wealth individuals. Everyone else you kind of go to one size fits most model.
Charley Ma (40:41):
I think now with LLMs it's still really, really early for all the downsides we talked about before. But I think we're going to start to be able to actually see the promise of personalization and an actual utilization of data.
Charley Ma (40:53):
But also, to you all's point turns out, yeah, junk in, junk out, when it comes to data, the most important thing is at the end of the day fixing your core infrastructure, making sure you're actually having access to a unified data stack that you could actually draw insights out from. And the AI is not going to solve that, that's just your core fundamental architecture, the providers that you're picking and how you're just running your banker FI.
Jim Marous (41:17):
Well, it gets down to listening. I brought this case study up a lot lately because it happened recently, is that I was trying to find a local Jeep dealer that I couldn't remember the name of that I'd been to before.
Jim Marous (41:31):
And I was asking my wife at our house and saying, geez, do you remember the guy's name? No. I look over her shoulder and see that on my Hulu screen, on my TV, in the middle of the news, they cut into a commercial for this one dealer. Well, that wasn't the dealer that I wanted to know about because I knew about that one.
Jim Marous (41:50)
The next commercial, immediately after it, was the dealer I wanted. And what that means is that the consumer's getting more and more used to being listened to.
Jim Marous (41:59):
I think Bank of America has an amazing lead on everybody when it comes to LLM because of all their experiences with Erica. We laughed at its beginning, say, well, that's cute, it's fun. But the reality is the amount of data they collected and the conversations that they had with their customers, a massive number of digital banking customers puts them in an enviable position when it comes to generative AI and being able to deploy against that.
Jim Marous (42:24):
Finally, Colton, from your perspective, again, you see a lot in the marketplace, what's the one recommendation you give financial institutions today to be able to innovate better?
Colton Pond (42:35):
I'll go to a lesson recently that I was talking with our co-founder and CEO Rhett, who started LoanPro 15 years ago. And I was talking to him about lessons he's learned over the past 15 years, and he started LoanPro very differently. LoanPro’s been bootstrapped, not venture capital funded, cashflow positive from the very first. And I was like, hey, what lessons have you learned? And one of those was, take strategic bets on innovation.
Colton Pond (43:06):
We were customer number 47 on AWS very early on, and we saw, we, not me. Rhett, and the founders saw that's an innovation that we believe that we can build on. And we'll be the next wave of how everything, and now today, everything's in the cloud. Everything's built on either Google or Microsoft or AWS in some capacity.
Colton Pond (43:30):
But I think the hardest thing to do, and the reason why Charley brought up like these forums of banks, learning from banks and credit unions learning from bank and fintechs learning from banks is important because you have to cut out reality from the noise. And that's the hardest thing to do. But when you see that and you see things that are working in other institutions, the early adapter move or entry is real and impactful.
Jim Marous (43:57):
Boy, that's so important in the banking world. We used to use the word fast follower. We were neither fast and we were always a follower. And the reality is, because everything's happening so fast, there's very little prize left if you're a fast follower, it's just a matter of keeping up.
[Music playing]
Jim Marous (44:11):
Charley, Colton, great conversation. I really appreciate your time today. I think we covered a whole lot of ground. Charley, good to meet you for the first time, I believe. And Colton, always good to be with you. So, thank you very much, both of you.
Colton Pond (44:23):
Thank you, Jim. Charley, thank you, my friend.
Charley Ma (44:25):
Yeah, no, thanks for having me.
Jim Marous (44:27):
Thanks for listening to Banking Transformed, the top podcast in retail banking and the winner of three international awards for podcast excellence.
Jim Marous (44:34):
If you enjoy what we're doing, please be sure to give us a review, a positive review hopefully, on either YouTube or on our podcast channels.
Jim Marous (44:43):
This has been a production of Evergreen Podcasts. A special thank you to our senior producer, Leah Haslage, audio and video producer Kae Holmberg and video producer Will Pritts.
Hide Transcript