Fair Isaac Corporation (FICO) Earnings Call Transcript & Summary

All right. Good afternoon, everyone, and thanks for joining us. My name is John Mazzoni. I am an equity analyst here at Wells Fargo. And for those of you in the room, we're at the Wells Fargo TMT Summit here in sunny California. We have the pleasure to host Steve Weber, the Executive Vice President and Chief Financial Officer of FICO or Fair Isaac Incorporation (sic) [ Fair Isaac Corporation ]. Steve leads FICO's global finance organization and he has been with the company since 2003. Previously, he has led both treasury, tax and investor relations departments. And during his 20-year career at FICO, he served as the head of FP&A and also has an intimately involved strategy. With more -- with less ado, do you want to maybe kick it off with a brief background and just what FICO really means to a TMT investor base?

Yes. Great, thank you. Thank you for inviting us today. It's great to be here. Yes, FICO has been around for more than 60 years. People think of us as a scores company. And obviously, that's a big piece of what we do. That's half of our business roughly, and it's a very profitable half of our business. But we also have a pretty significant software business. We started building software primarily for banks in the '80s, and we've done some acquisitions along the way. And today, we have a really interesting platform business. It's taken a lot of the IP that has been embedded in our software, and we have it available on a platform and it's growing really dramatically. We're seeing -- this last quarter, we had better than 50% ARR growth in that area. So it's really -- it's the right product for the right time, and banks are all talking about improving their digitization of their business, and we are really helping with that whole transformation. So it's an exciting time to be at FICO.

That's great color. Thank you. And maybe just to talk a little bit more about the scores side before we jump right into software. Could you just remind us on how the FICO Score originated? And really, what was the basis of the decisioning process? Maybe more of a historical lens, if you will?

Sure. Yes. So I mean, Fair Isaac was founded by 2 Stanford graduates whose premise was that if you take data, you can make better decisions. And so it really was one of the first analytics companies, and that really made its way into the credit decisioning process in the '80s when they built scorecards first to figure out how do we take a process that has been pretty subjective. If you go back far enough, you had to kind of know somebody to get a loan and to really take -- look at data and make objective decisions. And that's what they tried to do. And the first score was built on top of Equifax data back in the '80s, and then we built the score across all 3 bureaus. So we were able to take these credit bureau reports that had a lot of information in them and then boiled that down to a 3-digit number that can kind of identify what's -- how risky is this person. If you were to lend to them, what's the risk associated with them? So it really was one of the things that spurred a lot of nationwide lending. You didn't have to go into a bank and try to explain to them why you were creditworthy. Suddenly, they could take all the data they had about you and make decisions on you remotely.

Absolutely. And maybe just double-clicking into that lens. It seems like this -- the penetration, again, based on industry estimates, is above 90% and you have over 10 billion scores a year. How should we think about the evolution of the FICO Score and growth as well as potential for pricing to the value, also known as special pricing?

Right. So the first, probably, 10 to 20 years was all about gaining penetration and really seeing the score get used more and more. [ Philip ] became so ubiquitous that virtually every lending decision uses the FICO Score. It's used to help generate the loan in the first place, then it's used to -- we sell account management scores, which are used to help identify their ongoing risk in a portfolio and understand what the underlying risk that you have. And that can also be used for securitization. If you want to securitize the portfolio, this is the measure that they can use to determine how risky that portfolio is and then determine how to price it. So it really is used for a wide number of things, even marketing upfront, figuring out who to send offers to. So there's a lot of use cases for it throughout the credit ecosystem, in mortgage, in auto, in credit card lending. So it's really widely used. And for a long time, we never changed the price. So for the first almost 30 years, it was the same price that was put in place in the '80s. So in the last 5, 6 years, we've started to actually push through some price increase because we realized the value that was derived from this score. We weren't pricing anywhere near what that value was. So we've been putting through price increases every year to kind of bring us in line with what we think the value is.

Absolutely. And then maybe just quickly touch on the consumer because it seems like there's been a lot of jitters in the economic backdrop. Have you had any updates in terms of what you're seeing in terms of not only the prime and subprime consumer groups but perhaps by mortgage, card and auto as we should think about going into '24 and '25? And how is really the FICO Score perhaps more resilient than it was in the past? I'll leave it there.

Yes. Well, I mean, I think, first of all, the score was always predictive. I think one of the things you saw in the financial crisis is that lenders were purposely lending to subprime markets. And by definition, subprime means they're low scored. So really, what they weren't doing was they weren't doing a lot of their typical underwriting practices. They weren't checking for income verification. They weren't -- they had improper loan-to-value ratios. They were sometimes lending on houses more -- they were lending more than the value was. So I mean, it was just a problem waiting to happen. And then they had a lot of things with adjustable rate mortgages as well. So I mean, we -- what the FICO Score does is it identifies -- rank orders people and identifies who the subprime is, who the prime is, who the super-prime is. So the farther down you go, the riskier it's going to be so it's up to the bank to price it correctly. But what -- in terms of what we're seeing today, there's still a lot of strength among consumers. I mean, the subprime is getting squeezed certainly in terms of credit. It's a lot harder for the subprime consumer to get access to credit today than there has been. But there's a lot of lending activity that's happening, even on the credit card side, which tends to fluctuate a little bit. But there's a -- if you're a mid-market or an upmarket consumer, you'd have no trouble getting a credit card. A lot of what's happening in mortgage is really the market for [ OSB ] because of the rate increase. The dramatic climb in rates in such a short period of time, you have buyers that don't want to buy anymore and you have sellers that they don't want to sell because they got to buy something else in some cases, and then they don't want to lower their prices. So it's kind of froze the market until we get to some sort of equilibrium or if we start to see rates come down.

Absolutely. I think that segues nicely into the software side. Could you just again, building on your expertise as an analytics company and as a decisioning company, how does the software really fit into the bank workflows as well as the overall credit landscape?

Yes. So I mean, banks more and more are trying to -- again, the digital transformation is what they're all chasing today. And really, it's because the pandemic really kind of spurred this along. People don't go into branch offices much anymore. In some cases, there are -- branch offices are closing down. So the one-to-one interaction doesn't happen anymore in the same way. It used to be you go into the bank. You'd sit down with somebody there that would kind of walk you through what -- if you're interested in this or this, you talk to that person or whatever. That doesn't happen anymore. So the banks need to be more responsive and proactive, frankly, in how they deal with their consumers. And they have a lot of data on their consumers so they know a lot about the consumers. In some cases, they have -- if the consumer has a DDA account there, you know what their direct deposit is every 2 weeks. You know when they got a raise. You know because you have access to their credit [ bureau ] that you know when they paid out their last car loan. You know a lot about them. And knowing all that and being able to analyze that on an individual person basis gives you the ability, if you apply analytics on top of that, to make the right decision moving forward with them. Every interaction has to be optimized. So if you know the person, maybe they got married. There's suddenly 2 people getting direct deposits now. So maybe now is the time to offer them a car loan or a mortgage or whatever. So it's all those things that you realize what a person is likely to want next based on all the aspects you see of them, you can make better suggestions to the consumer. And in a proactive way, rather than wait for the consumer to come to you because the consumer might not come to you. They might go to somebody else. But if you can be, through your app or through e-mails or voice mails or whatever, if you could be better at interacting with that consumer, the effectiveness is really high, and a lot of banks are starting to realize that today.

Absolutely. And what are the main use cases that FICO -- again, some people in the room are newer to the story, the main use cases that FICO really solves for the banks and also serves within the entire life cycle, be it fraud or consumer communications and others?

Right. So historically, we had point solutions that would deal with originating a new account, and then it would hand it over to somebody in account management that will help manage the portfolio. Who should get line increases? When should we change interest rates? We had a collections product that would help deal with collections for late -- if there was late payments. We had the Falcon Fraud Manager that looks for credit and debit card fraud. Again, we still have that, obviously, and it's a -- it has like 2/3 market share in the world. So it's a very highly used product, but they were all point solutions in specific silos within the bank. More and more banks are trying to look at across the entire bank and look holistically at the customer and figure out what do they need. And by sharing all the data from all these different areas and applying analytics on top of that, you can make better decisions on those interactions with consumers. And that's what the platform is all about. That's the difference between the old legacy products that were siloed and that were point solutions and the platform, which kind of breaks down those silos and allows the flow of data back and forth.

Great. That actually is what my next question was. So the platform has had over 40% ARR growth for about the last 14, 15 quarters. And right now, we're seeing 55 banks within the enterprise kind of platform customer program, which means they're using more than one solution. Could you just talk about the land-and-expand strategy and really continued adoption? How much runway there is left?

Yes. So we have a little over 100 customers that use at least one use case. And then we have 55 that are EPCs, your enterprise platform customers, that use 2 or more use cases. Because we figured if they're using one use case, they really aren't a platform customer yet. They really are just a point solution customer. But the strategy is to get the platform installed and get all the plumbing done, all the data hookups ready, so that the customer can see what's possible and how easily it can be used and the flexibility that it has. And when you get it started, when you get it installed and implemented, invariably, the customers grow dramatically. So we've tried to drive down the initial expense, try to make the PS, the professional services lift, as light as possible. So there's not as much expense to get it installed. And then, again, trading them up as much as possible on what's possible. And we have a customer success team that comes in behind the original sale that tells them what other banks are doing and saying, well, you could try it this way or in this area. And like I said, invariably, they'll increase. So it really is a land and expand. I mean, our ARR growth is 50% or better but our net retention rate is like 140%. So I mean a lot of it is coming from you install it at first, and then you ramp them up afterwards. So there's a lot of runway left. I mean, we've got $170 million, $180 million of ARR today on the platform. And a lot of industry analysts looks at this kind of the environment here, depending on how you measure it, but it's probably billions of dollars a year, tens of billions of dollars a year in potential revenue that's out there in total. So we're well on the way to at least getting a foothold in this space. And we're way ahead of anybody else in terms of serving this market because it is a fairly new market.

Absolutely. And I think no matter how you cut it, there's billions of dollars in TAM and that's mostly just due to the fact that these are not only U.S. banks but global banks.

Right. It's definitely a global marketplace because a lot of our early wins came from outside the U.S., Canada and Brazil, in particular. In the last year, we started to see a lot more wins in the U.S. and they're just kind of early stage at this point.

Definitely exciting. And maybe just segueing into the investment, you guided to about mid- to high single-digit expense growth within the software side. Where do you see the major buckets of investment? And how are you balancing that with margins on a longer-term basis?

Yes. It's R&D for the most part. I mean, a lot of it is -- again, we have this platform that works really well, but we need to build some efficiencies into it today so that we can enjoy those as we scale the business. So we need to do as much as we can to be able to run multi-tenant as much as possible. We need to be able to simplify some of the install, so there's less hours that are required to install. We do have less of our own professional services resources. Or as we look to move to work with system integrators, we want to make it easier for them to do as well. So we're building out APIs for them to use. We're building more documentation out so that they can actually install it themselves. As it stands today, almost every implementation requires our own in-house professional services people. And we want to get away from that.

Got it. And maybe just to try to get a sneak peek of FICO 2024. Are there any exciting product announcements or any potential things that we...

There will be a lot of exciting things that we can't -- I mean, it's 5 months away, of course, something like that. It's in April, right down the road in San Diego. But there's a -- we've got a big event planned, and it's going to be really exciting. There's a lot of things that we'll be rolling out. There's a lot more function. I mean, you were at the last one. You saw a lot of what we did there. And every year, it gets a little bit easier in some ways because there's a lot more use cases. And there's a lot more customers that are doing really exciting things that they want to talk about. So we're looking forward to a big event this year.

Absolutely. And maybe just also hitting on FICO World. Do the banks have very similar problems? It does seem like there's a lot of cross-pollination, and there's best practices being shared at FICO World. How does that segue into maybe a contract signing or a potential conversation or a door opening?

Yes, we have a lot of really good conversations there because a lot of the banks are coming there. There's a lot of cross-pollination, but then there's also a lot of -- there's also -- they kind of keep their cards close to their vest on some things, right? I mean, there's -- everybody has kind of their own proprietary secret sauce that they want to do and what -- the great thing about the FICO platform is that you can build -- we're giving them the tools really. We're giving them the tools to use. If they've got a lot of data scientists that can glean their own insights, we're giving them the tools to do that. We're giving them some analytics, some modeling capabilities and the ability to put this into a reproducible workflow that this is the decisioning behind the workflow. So there is a lot of cross-pollination, but there's also still a lot of them are owned that they kind of want to keep to themselves. So they're willing to give up some to their peers. But in other cases too, there's a lot of things that are codenamed.

That makes sense, the delicate balance. And maybe just going a little bit further into the platform itself. It's not limited to financial services, but you've seen optimizer use cases within companies like DoorDash. You've seen explainable AI. Can we just talk about the longer-term runway of the platform and really how to think about it as a stand-alone asset within kind of the marketplace?

I mean, yes, taken at the highest level, it really is a set of capabilities and technologies that can ingest a lot of data, make decisions on that data, and then put that decision into a workflow. So it really works best with consumer information. That's kind of what we -- that's our background is in consumer. So we're really good at taking massive amounts of data on consumers and then taking all the learnings from that and making decisions at the individual level. So you really do have a one-to-one decisioning. It's not -- in the old days, it was always crude segmentation. You put everybody into one of 5 buckets. They were a soccer mom or a young professional, just the basic things like that. And that's fine, but you can do so much better now because you know so much about the individual that you can make decisions on that individual. And rather than just send out an e-mail to everybody saying, hey, we have auto rates as low as 5.5%, you send something to that person with a specific about we can lend you this for this many months at this rate, prequalified. And you say it to them because you think they might want that, right, not just because out of the blue. You've done enough research that you think this is what they're going to respond to because more and more, it's the kind of thing where people get so much information, so many e-mails. They're flooded with so much. You have to make sure that whatever interaction you have is meaningful. Otherwise, you're not a trusted vendor to the consumer.

Absolutely. So digital personalization at scale.

Exactly. It is at scale. And everybody needs that. It's not just financial services that are used to making decisions on data, obviously, and they have the highest cost decisions in a lot of cases to make. But if you drive the price down, it's available to a lot of people. So we do a lot of work in insurance. We do a lot of work in telecom. We do some work in retail. There's a lot of different areas that others could use this. And as we get away from financial services, we're probably going to rely more on system integrators because there's a lot of them that are looking for more IP to build practices around. We've got IP that's growing really dramatically, so there's obviously a desire for it. And they've got a lot of contacts and a lot of domain expertise that we don't have. So we have a lot more capacity than we will ever be able to distribute. So I think there's a really good opportunity for us there. But that's probably down the road probably. That won't -- we'll start to maybe announce some of those deals in '24, but it's '25 and '26 before you actually see revenues.

So gauging much more runway for growth. And also just one quick point. There's also the side of the fraud. Have we seen an uptick in fraud? Just because looking at not only kind of the bank workflow itself but also the consumer protection side, I think that you have a very strong market position within fraud as well as customer communications and account management. But when we think about just the non-platform side that sometimes gets overlooked, it doesn't seem like you're decommissioning these products but you still are kind of doing the blocking and tackling.

Right. So a lot of the legacy products that are off-platform for us, they're deeply embedded, they work really, really well. And they -- like our Falcon Fraud Manager, it's typically run on-prem at a bank on the bigger banks or it's run through a processor for the mid-market banks. And it runs potentially billions of transactions a day. If you think about if you're a large bank, somebody is using one of your credit cards all the time, 24/7, somewhere in the world. So you can't really just take it down to move to a different solution or even sometimes, upgrades are difficult. And it works extraordinarily well. And from our point of view, it's profitable for us. It works well. The banks are happy with it. So it will be a long time before those actually go away onto something new. We have functionality in the platform that we're rolling out to kind of take some of it on. And over time, we'll probably see that happen. But a lot of those products, they just really work really, really well, and we don't see them going away anytime soon.

Maybe not anytime soon. But if we take a longer-term lens, is there a potential for everything to move to the platform?

Absolutely. But I think it's a long time.

A long time.

I mean I think there's a lot of things that are going to move the platform before those do. I mean, even from a bank's point of view, there's a lot of things that they can do right today on the platform that are pretty easy to get started on the platform and that they don't already have a product for, right? So there's a lot of things. I mean, Falcon works really well, but there's a lot of other things they're trying to do that they don't have a product for it today. So if they can get that done in the platform, that's what they want to do first.

Absolutely. And there also might be regulatory requirements and data requirements, and you work globally with different customers.

Right, right. And that's always been a challenge. I mean, working globally, for a long time, people thought that banks would never move to the cloud because of security or they'd never let the data leave their 4 walls. And there's still a lot of regulation around what data can leave the country. A lot of data needs to stay in country. So for a while, we had to do -- we had to have our own data centers in a lot of countries until AWS was ready for that. But now we see -- we use AWS for almost everything. We're trying to decommission our own data centers as much as possible because obviously, AWS can scale data centers better than we can. So we're happy to work with them, and they have such a scope worldwide. So that's been good for us.

Great. Maybe just shifting gears now to capital allocation. FICO has been, as of late, a proponent of buying back their own stock as opposed to M&A. And given...

It's been more than as of late. I mean, we've been like that for the last probably 15 years or so.

A decade or so, give or take. So how should we think about capital allocation at a high level? As well as kind of given how much free cash you kind of blow off at the business model itself is inherently a machine and with kind of a target leverage ratio, that 2.5 to 3x rule of thumb, you fit firmly in there? Because looking forward, where do we see kind of capital allocation evolving?

Yes, it's a great question and we get it a lot. I mean, I think we always look at different M&A opportunities. But it's hard to find something that doesn't dilute the business, frankly. It might be fiscally accretive, but we like the business as it is now. We've got the 2 different businesses, the scores business and the software business. Theoretically, there's a lot of software assets we could look at, but we like the platform we have. We don't really want to go out and buy another product set that has different architecture that we have to either integrate somehow or run it as a separate business. That doesn't look very attractive to us. So occasionally, we'll do things where we will buy technology that we can plug onto the platform pretty easily, but those tend to be pretty cheap. They don't cost much. On the scores side, we haven't done a lot there. There's really not a lot of -- I think we can do some things internationally potentially, but we like the scores business the way it is. We like the software business the way it is. And we -- our first call for dollars is to invest in those, right? So we make investment every year on those. And then the excess cash flow, we tend to either buy or pay down debt or buy shares back. We don't have a dividend, but we like buying back the stock. We think if -- it's worked really well for us. If we can grow our top line double digit, do a little bit better than that on net income and at the same time take out shares, you get even more benefits. So that's kind of our formula, if you will. And with rates being higher, we tend to look at the cost of the debt a little bit more. But still I mean, we -- this last year, we saw our leverage tick down a little bit because we didn't borrow more to buy shares back. So our EBITDA keeps growing, so there's a lot of options for us. But we like the buyback. We like -- we're -- our CEO, Will Lansing, jokes that we're in a slow-mo IPO, right, that we're just taking them out one at a time. And at some point, everybody will fit in this room, the shareholders, but it gets harder.

I was trying to avoid the slow-moving IPO characterization, but I think it's a high-class problem to have and also beg the question of what is the competition like on the software side? And really, who are the main competitors? Are they the banks? Are they external? Is it potentially an in-house solution?

Yes. I mean, a lot of what we're displacing is in-house solutions. And a lot of that tends to be pretty manual. It's -- a lot of it's teams of people that work on something. They'll build a model and then they'll take it down to engineering, and that gets coded in and then they track it for a while. And then when they want to make changes to it, they build a new model, then they take it down to engineering, get that coded in. So this kind of displaces a lot of that. It gives them the opportunity to be more -- to do more tests and to run different scenarios and simulations and see how it works in the test environment. And then they can put it themselves into production without having to go to engineering, and then they can track it and they can make updates more frequently. So that's kind of what it's replacing. In terms of competition, it really comes down, for the most part, do they want to go with point solutions or do they want to go to a platform? Because from a platform point of view, we don't really have any competitors in this space today because it is fairly early stage.

Got it. And is there any inherent resistance? Or maybe, could you frame the roadblocks you're facing within those banks? Is there potentially an in-house point solution that might be trusted and true, but how could that maybe not be future-proofed?

I think there's probably less pushback. I mean, any time you have new technology, there's always pushback. There's always a concern that, does this work as well as we think it does? What are the problem -- I mean how long it's going to take to implement? And how much it's going to cost? And what do we do if it doesn't work? So those are always concerns. I think there's a lot less of that today. I mean, it's a different way of doing business. You're trusting technology more, certainly. But I think they realize they have to. And again, the days of conducting all your business in a branch office are gone and they're not coming back. So more has to be done through an app or through the website or through whatever method you have. And in order to do that, you have to have decisioning behind it, right? So you can't just go to an app and say, okay, I'd like to apply for a loan and then get back to me in 2 weeks, right? Those days are gone too, right? You have to have decisioning built into this that can kind of keep working you through the process and that can even make the decision.

Great color. And maybe just it would be remiss not to mention the regulatory landscape. How are the regulators responding to this? How are they changing? And what has their feedback been thus far?

On the software side, I think it's all good. I mean, they're regulators. First and foremost, they're concerned with risk, anything that can take a risk out of the system. And they're also concerned with fair lending, right, and access to credit. So really, we spend a lot of time on all of those, right? I mean, we're probably the leader in a lot of areas of risk. Most of what we do is around risk and risk management and helping regulators understand how much risk is this bank taking. So we're a key part in the process as that risk is evaluated. And we're doing a lot on financial inclusion. We're looking for ways to bring in additional data sets to help the under-banked because a lot of the -- a lot of people have been kind of shut out of traditional markets because there's not much information available on them. But as we can pull in things like rental payment data or cell phone payment data or utility payment data, those are all things that are certainly predictive but are not typically in a credit bureau report.

Absolutely. And that speaks again to the kind of nature of the FICO Score and the dependability on the benchmark status. Maybe could you just remind people in the room about VantageScore and perhaps some of the most recent developments with FHFA most recently pushing out the bi-merge decision from tri-merge? And really, FICO's approach to that and any potential regulatory/landscape issues that could arise from it?

Yes. I mean it's hard to know what's going to happen and when because -- so just for context, the FHFA sets the rules by which conforming mortgages will be accepted. And for a long time, they said you had to have all 3 credit bureau reports, Experian, TransUnion, and Equifax, and then you have to have a FICO Score on top of each of those. What they did about a year ago now, a little over a year ago, they came out and said that at some point when this is implemented, you're only going to have to take 2 of those bureaus. You can take 2 bureau reports, and then you'll add there 2 scores on top of those, one VantageScore and one FICO Score. They laid out a time line a year ago and they keep pushing it back. So they're getting more -- they're asking for more response from the industry in terms of what the industry is looking for, concerns that might have. And the industry is coming back with some different concerns and pushbacks. So it looks like it's going to be a few more years before any of this gets resolved and implemented. But it's good that they're taking their time because this is a pretty big shift for the industry, and you obviously don't want to change anything without thought, thinking through all the different permutations.

Absolutely. And at most banks, there is going to be a chief or a C-suite or Board-level approval at some, when you change risk models, when you change different decisioning. And again, just going into perhaps the shift from kind of the FHFA decision, could you just, again, frame what the kind of complications were with kind of FICO 10T as well as VantageScore? And is there any potential kind of longer-term ramifications? Or do you think this was a justification of why FICO Score is as needed as always?

Yes, I think FICO Scores is needed as always. I think they went through a long process to determine what to do. I think they figured they had to do something, so I think this is what they did. And they took a good faith swing at this whole thing, right, and tried to determine what to do next. And I guess we'll see when it gets implemented how it -- but it's difficult. It's really difficult because you're dealing with the entire mortgage industry in the United States.

Absolutely. And maybe just one or 2 more high-level questions. We're almost up on time. As you look at the business today and take a longer-term view, especially with people who are more focused on the technology side, could there be a potential risk from gen AI or an opportunity from gen AI? Or is it less applicable in this space?

I think -- well, separating the 2 off. On the scores side, it's not very applicable. It's really difficult to use AI for credit decisioning because the Fair Credit Reporting Act is such that you have to make sure you don't have disparate impact against protected classes. You have to make sure you can explain the decision you made. If someone applies for a credit and gets turned down, you have to tell them why they were turned down. There's a lot of steps along the way to protect the consumer. And then on the risk side, regulators demand that any time you make changes to your model is you explain that risk and you understand the risk associated with that, what are the changes, why did you make them, what's the impact going to be. And it's difficult to do that with AI, right, because AI is not built for that. AI is built to just continually update and generate better and better results. But if you go down that path, it's difficult to meet the requirements of the law from a regulatory environment and still get the benefit of AI. Where we are using AI is much more on the software side, like things like fraud. We've used AI for more than 20 years on the fraud detection side. Like Falcon has AI built into it to help improve the model, help find the fraud that's out there, and then continually update the models. Yes. And then beyond that, we're using generative AI in things like coding. Sometimes, we do it for idea generation around different things. But typically then, it gets passed over to a human to push the process along.

Absolutely. And not even to mention the risks of hallucination and bias and other things that really cannot be tolerated under that regulatory structure, which...

That's what happened. I mean, that's what you have -- that's what ends up happening. You basically get digital redlining.

Digital redlining, of course. Again, anything else that you would like to mention to investors today or how FICO should be thought about? Because again, it was seen as more of a FICO Score and now that you have a software business that's 50% of your revenues and growing at a very healthy clip, especially on the platform, there's a lot to be excited about. What should we take away from this conversation? And what should we think about for the future of FICO, give it 10 years and plus [ a few ]?

Yes. I think the business has changed dramatically in the last -- I mean our -- Will Lansing came in, I think, 12 years ago. Dramatically changed the business, right? So look at the software business as completely a different way in terms of taking really legacy products and putting them in the cloud first and then building a platform that can really grow because we were a low single-digit grower on the software side. There wasn't much room for growth there. And on the scores side, looking at that business and realizing that there's a lot of untapped value there, and we need to somehow capture that value. And it really has changed the business completely. And we now have 2 really strong segments that are both performing very, very well. Very different. They have the same customer bases, but they're very different. So I mean, from the outside, it's hard to kind of figure that out. If you're a specialist, you might not know much about software, but you do with business services and vice versa. So it can be difficult from that point of view. But it's an exciting time because we have a lot going on, and we're really starting to see -- the last several years, we've seen a lot of growth in the scores side. But now we're really starting to see the software business, what that can become. And there's a lot of -- we're not focused on near-term margin expansion, but there's going to be a lot of margin expansion to be had there as that business scales.

And just to characterize, would that be more of a 5- to 10-year time frame?

I think it's closer than that. I think it's -- I mean, we saw some in the last few years. We've seen some expansion. There's some reinvestment we're doing next year. But I think in 3 to 5 years, you're going to see some decent margin expansion. A lot of it depends on what the growth trajectory looks like. If it's growing fast, we'll keep reinvesting. But this -- I mean, we could drive margin expansion really quickly, but that would have an impact on the growth. So we always look at that. But there's -- we're beginning to see even in the last couple of years pretty significant margin expansion for us.

Absolutely. And then with the final question, I think it's a perfect segue. Just is there a reason in the longer term that the 2 businesses need to be combined? Or could they stand alone as a scores business and a software business? Because it sounds like the opportunity for the software side is so large at such an early stage that there could be a reason to maybe potentially spin or sell. How do you think about that at a high level?

It's certainly -- they don't have to be together. I think there's a lot of advantages to having them together. I think in some cases, having them together has given us a little more cover to do some of the harder decisions that would have been investing in some of these things. But they don't have to be. I mean, if there comes a point in time where the consolidation discounts too much, we can certainly split the 2 if we wanted to. But I don't think -- the time is not there yet. I think there's a lot of value that's going to be seen on the software side in the next 2 to 3 years, and I'd like to have the current shareholders get the benefit of that.

That's great color. Thank you again for taking the time today, and we'd like to host you here at Newport, California. Sorry.

Great. Thanks a lot.

Thank you.