Fair Isaac Corporation (FICO) Earnings Call Transcript & Summary

All right. Welcome, everybody, to the fireside discussion for FICO. We're super excited to have Will Lansing with us. You're somewhat fresh off the plane from Montana.

That's right. Just now...

We're just talking about the fact that the company got added to the S&P 500 Index effective or so, right, the 20th or something. Not a lot of S&P 500 companies headquartered in Montana.

Yes, it's true. It's true. We might be the only one. I'm not sure...

Yes. There's Snowflake.

Sure.

Okay. So there's 2 out of 500. Okay. And Bozeman also...

Yes.

How did it end up there?

It's a long story. But part of it, I live there. So I'm that CEO, who brought the headquarters to where I live. But it's also -- we had a great little office there. And it was good.

But the footprint is not -- head count is not concentrated there.

We're really a super virtual company. So we have 30 people out of 3,300 in Bozeman, but really our people are scattered all over the world. Even before COVID, I would say, nearly half of our people were home officed because we go and try to get the talent wherever it lives. And so we've never had a lot of pressure to be in the office.

Right. So return to office, not a big dynamic.

We have -- I mean, there's one place. Our Bangalore office is about 1,200 people, and that's very -- there's a culture there that's -- there's a lot of training that goes on there. And so that's one place where we really do have that kind of office culture. But I'd say outside of that, we have offices scattered all over the world, but they're lightly used.

Yes. Okay which then creates some efficiencies opportunities and...

For sure. We've taken out probably 20% or 30% of our real estate.

Yes. Cool. All right. Well, so we -- I think we're talking about this last night. In a sense, FICO was doing AI before AI was as hot as it is now. What does AI mean in your context?

Well, so just take a step back. FICO has been an analytics company since the 1950s. And so we've kind of grown up, trying to leverage every analytic technique available to make data-driven decisions. And so that's gone on 50s -- up 70s, 80s and then we kind of got into software. It used to be more of a consulting analytic assignment and then we got into software. We've been doing neural network. We've been doing AI for a long time. I think that the real question is, to what end? What are you going to use it for? This isn't like a hobby. This isn't a theoretical exercise. You have to have a point of view about what you're using it for. And that's where we have -- our deep experience in taking data-driven decisions and getting them into workflows and getting them into that point of contact with end consumer. That's where it really matters. And so when it comes to AI, we're very focused on applied AI, Applied Intelligence. Can you make a decision and then get it into a workflow so that it can be used. And then -- and a corollary to that is explainable AI. So we operate in a pretty heavily regulated sector, financial services, and it's not okay to just make a decision and not know how you got there. It's critical that you can explain to regulators so that you're in compliance with the law, with fair lending laws. It's critical to be able to explain how you got there. And so we have this notion of ethical AI, explainable AI, and we're very focused on being able to explain how we got to the endpoint.

Right. So explainable, responsible, ethical. Is there a difference from one jurisdiction to the other that's notable, one government versus another?

I wouldn't say so. I mean, I think Europe is a little bit ahead of us on kind of -- [ numerous ] rights to the data ahead of the U.S., I should say, on consumerization of the data. But we're focused on all of that.

So good relationship with the regulators because, I mean, a lot of the big tech companies have a fraught relationship.

For many, many -- we're not a regulated entity like our customers are. But because the regulators rely on FICO scores, because they want to understand how the software works, we've been very focused on what the regulators want for a very long time. And our customers, the banks have come to view us as a partner in explaining to the regulators why and how things operate, what the risk level is portfolio by portfolio. And so yes, we have a lot of -- we pay a lot of attention to that.

All right. So what's your crystal ball tell you about the direction of the economy and the state of the consumer in the U.S. and all the other important places?

I guess I would qualify that by saying that we get our data in arrears. We are 6 weeks behind the credit bureaus. But what we're seeing is pretty much what we expected, which is volume decline, but slowing rate of decline and obviously, it varies by submarket. But scores volumes are lower than they were a year ago. We've made up for some of that with price. And I think the consumer is actually pretty strong -- and we're seeing pretty good strength in credit card activity. Auto is stabilizing. Mortgage is down, but that's not a shock to anyone. That was expected.

How about -- like where some of the fintech innovation has happened over the last cycle, like I'm thinking buy now pay later, right? It went up and went down, but that was some "innovation", are there things like that out there that you think over time drive more volume?

Yes. For sure. So one of the big challenges for us and for everybody in the industry for all of our customers is how do we evaluate more people, get more people into the system, score, and then provide credit to people who are otherwise credit invisible. And so we spent a huge amount of time trying to score additional population. And we have a lot of innovation around that. We have scores, new scores that are designed to capture responsible behavior in populations that aren't normally captured by the credit bureau file .

And then you take some on pricing. So this idea about getting paid for the value deliver been part of the operating philosophy. But how do you make sure you don't overfish the pond on pricing?

Well, so maybe a little bit of history is worthwhile here. We've been providing scores. Let's talk about how this full score thing happened. Where did FICO scores come from? How did we become an industry standard? If you go back in time to the 50s and 60s and 70s, we would build proprietary scores for bank customers. And we say, here's a scorecard, here's how to evaluate people, low-cost way of doing it.

Mainly U.S. banks?

In the U.S., primarily in the U.S., although we now have international scores also. In the '80s, we had this idea. We partnered with Equifax and the -- we call the Beacon Score. And we launched this score that would be available to any bank that wanted to buy it built on the Equifax credit file. And that was a big hit. That was like suddenly, you didn't have to be a bank contracting with FICO to build you a proprietary score card. You just go buy the score from Equifax. And so that was pretty good. Then the next innovation was to provide a score with the same [ opposite ] score ratio for TransUnion data for Experian data. And so we did that. And of course, the lenders love that because I wouldn't say that it commoditized the bureau file, but it gave them a little more leverage and a little more flexibility. And so they really liked it. And then the regulators saw that and said, "Wow, this is a pretty good metric for understanding the risk profile of the banks that we regulate." So they got into the habit of asking what's the FICO score for this portfolio, that portfolio, and how will it behave in the downturn. And then, of course, the lenders, they want to securitize the -- not always, but they often want to securitize the paper. And the investors say, "How do I evaluate the risk in this paper," and the FICO score is a pretty good way to do that. So then we have them in the game. And then I'd say the last innovation, innovation on our part to get deeply, deeply embedded in the system was around getting the consumer to demand it. So 10 years ago, FICO's unaided brand -- FICO's aided brand awareness 10 years ago was about 30% in the U.S. Today, it's over 90%. And some of that's attributable to marketing efforts by our partner, Experian, which has done a tremendous job of bringing up the FICO brand in the public's imagination. But I would also say a chunk of it was our Open Access program. where we said to lenders, if you're interested in sharing the FICO score, you're buying a FICO score for your own evaluation purposes. If you're interested in sharing that score with the consumer, you have the right to reuse it for free. And we thought that was pretty clever. We would really get deeply embedded as consumers started to see their FICO score every month. Well, the lenders kind of anticipated that and said, "Well, great idea FICO, but no, thank you." And then we had a couple of innovative lenders like Discover who said, "this is a great idea. Let's do it." And so they went and put the FICO score on the statements monthly on the website. And then many, many other banks follow suit. Today, we have 250 million accounts that see their FICO score monthly. So again, deeply, deeply embedded.

How much room is there for more innovation in the scores.

Well, a ton, a ton. So the value gap is huge. We price our scores from basis points to single-digit dollars. And typically, they're used in a decision that's worth hundreds or thousands or tens of thousands or hundreds of thousands of dollars. And so the -- and the score is probably the single best predictor of propensity to repay debt that there is. And so you'll obviously build other things around it, you'll check income, you'll do other things. But if you had to rely on one thing, it would be the FICO score. And so there's tremendous value there. We charge just a little tiny bit for it. There's lots of room for, over time, bringing up the prices to closer to the value. But there's also innovation. To your point, we don't want to just raise price. We want to give more value. And so for example, right now, we have -- in addition to the FICO scores that you're all familiar with, we've introduced a score called FICO Resiliency Index. And what that does is it lets the lender look at to 680s and decide which of these 680s is likely to perform well in a downturn versus the one that's going to get hurt more in a downturn. And that's pretty valuable because the blunt instrument in the downturn is for a lender to just say, let's just raise the cutoffs. We'll just go up to 700, we won't lend below 700. But if they're 680s, 690s, that are pretty good, it would be nice to know who they are. Well, we developed a score that lets the lenders do that. That's one example, but that's -- it's an innovation that we put in. If you buy FICO scores, you get the FICO Resilience Index with it.

So the predictive power can just get better and better.

Yes.

What's this concept of for permission data and what's the opportunity there?

I think that's a big, big, big idea. There's 2 pieces to it. One piece of it is there's a lot that the system doesn't know about you. And so the ability to score an individual is only as good as the data we capture in the bureau file and what we do with it. But there's other kinds of responsible behavior that, if you could get it into the decisioning, it would benefit a lot of consumers. And so you see that with rental data. You see it with telco payment data, cable payment data. And so those things are creeping into the credit file, where we're trying to do that. But with consumer permission data -- and it comes in 2 ways. It comes with Experian Boost, which is a way of boosting your FICO score and also with UltraFICO, which is another score that uses consumer permission data. What we say is, look, we don't know you that well. Come and tell us little bit more about yourselves and we'll see if we can't improve your scores result. And so we'll look into your checking account. We'll see you've had no overdrafts in the last 24 months. Things like that can actually help your score.

Got it. And then how about like the opportunity to build more of a D2C relationship between the company and the consumer versus the partners doing it.

That's a great question. And believe me, we've wrestled with that one for 10 years. So the question really is you have a consumer brands like FICO, very powerful. Consumers know it and value it. And how do you maximize the value of that to our shareholders. I mean, that's the question we wrestle with. And today, where we come out on that is there are consumer marketers who are bigger and stronger and better than we are and have more resource to promote the brand than we have. Experian, in particular, Experian is our big partner. We have other partners, too, but Experian is really our big partner. And they do an incredible job of leveraging the FICO brand for customer acquisition, for customer relationship, and we have a rev share with them. We have a super strategic relationship with them. And frankly, I don't think we could do what they do. Now we do have our own direct-to-consumer business, it's called myFICO. It's about $100 million business. It's an outstanding business and it grows and it's profitable. But I don't think we have the resources from a marketing spend standpoint to try to compete with our channel partners.

Right. And then there's some element of coopetition, [ tangent ] score [ as an example ], who do you have to balance.

Yes.

And as a practical matter with the directive to -- for mortgage lenders to use both and also by merge and trimmers like what are the practical outcomes?

Well, as a practical matter, we've been -- for those who don't know, in 99% of the scores purchases in the U.S., it's completely free market and the lender decides what score they want to use, and they almost always choose FICO. In the mortgage space, Fannie and Freddie, if you want to securitize your mortgage, sell it to Fannie and Freddie, they demand a FICO score. They want to know what's the risk on the paper. And recently, an evaluation process was undertaken in which should FICO continue to have this privileged position. And in the end, where the FHFA came out was we're going to have 2 scores. 2 scores are required, the FICO Score and the managed score. And so as a practical matter, it doesn't really affect FICO at all. There's the continued need for a FICO score. My own view is that investors downstream will continue to demand a FICO score, all of their models are built around it, and it's still the single most efficient way for them to evaluate the quality of the paper they're buying. So I think it has very little effect.

Okay, switch to software in this...

Sure. Let's [ go to ] software.

That's the upside, right?

Yes.

The bigger upside.

It is. It is.

Yes. Well, so there are a lot of moving parts, right? You've got the so-called established business, which is stable, and then the high-growth platform business. And there's lots of moving parts there. There's accounting transition, there's ramping up. There's lapping divestitures, the macro that needs to stabilize. But anyway, when all that is normalized -- not guidance, but like what do you think is possible in terms of long-term trajectory for software total?

The software business could be much bigger than the Scores business in the fullness of time. The TAM is enormous, maybe worth spending just a minute on the history here. So the history is -- I said we started as an analytics consulting shop in the 50s. We had this idea that if we could get analytics into software, not an easy thing to do, but analytics into software, we can get returns to scale, we could make money while we sleep. And so in the '70s and '80s, we started to build our own software, analytics software, which was adopted by banks, TRIAD and those kinds of -- and we also bought companies. We bought other companies. Most notably, we bought the Falcon Fraud, credit card fraud detection product, a company called HNC. And that was -- that is our legacy software business, it's about 5 critical applications for banks: Fraud; collected -- Collections and Recovery. We divested that, but fraud, and origination, and customer management. We figured out that we could get some returns to scale if we had a common decision engine across these different franchises. So we started to put a lot of energy into standardizing the decisioning within these products. within these offerings. That went pretty well. And then we took the next step, which was all these products, all these offerings require the same kind of data ingestion, data wrangling, data cleansing, data orchestration. So let's get the whole piping part of this to be more standard. So we did that. And it didn't take very long for us to figure out, this is now 2014, 2015, that we're really building a platform. We're really are building a decisioning platform that's more flexible and has more utility than just these 5 stand-alone applications, which are very important and have good market share, but they have very specific purposes. So over the last 7, 8 years, we've been on this journey to flesh out, put more features and functionality into the platform, what we call the FICO decision platform. And over the last 2 years, it has really captured the imagination of the financial services industry. So all the major -- I shouldn't say all, major banks who are now challenged with how do I do a digital transformation. How do I have a 360-degree view of the customer? How do we make sure that all my interactions across, all my channels are done with a view to what the other channel did with this customer, not long ago. All of that has put pressure on these financial institutions to transform the way they do business. Our decisioning platform is the right tool for the job. It's basically next-generation CRM. It's a decision platform that ingests any and all data from any and all sources, apply analytics, make a decision, feed that decision into a workflow such that when a B2C company for us, typically a bank, but any B2C company could do this, a B2C company that's interacting with its consumer in the call center, e-mail, text, walk into a branch or a store, whatever it is, all that's done with a view of the entire transaction history with that individual consumer. And what we do, we're in the optimization business. So we optimize that interaction for some objective function. Is it conversion? We want the revenue right now. Is it build lifetime value? Let's kick them into a more valuable category, even if it has lower conversion. Is it profitability? What are we trying to achieve? And then we can make sure that the decision engine drives us to the thing that we're trying to optimize. And that's really -- that's kind of what the platform business is. So we have this -- we have a half dozen legacy products that are really well established, deeply embedded, still on a very strong renewal cycle. And then we have a platform business, which is about $100 million, growing, 50% -- 40% to 50% a year and it is kind of the next generation of CRM.

Right. So everybody in the world is having to get the profit growth equation balanced correctly. I guess you could say that platform business gets subsidized by the legacy and by scores, if you want to look at it that way.

Correct.

But like how do you think about the investment priorities for the business? And what's the right cadence of that so that you don't go overboard and spend too much and get the right profit equation but still maximize long-term opportunity?

So because we have, the super profitable Scores business, and we have this high-growth software platform business, we do balance those 2. And every year, we think about it. But we do start with what does the software business need to continue its strong trajectory. And that's where we start. And it's -- we're -- although we're proud of the fact that our software business overall is now breakeven. It used to run $40 million, $50 million a year negative. It's now a breakeven business. And we like that, and the margins will improve as we get more scale. We're not so focused on margins in the software business right now. We're -- it's a land grab. We are clearly ahead of all of our competitors on this decisioning, next-generation CRM platform. The potential -- you said, what's the value? What's the potential? I mean I think if you did a sum of the parts on FICO market cap, we're $17.5 billion market cap today, $17 billion of that is our Scores business. Just -- if you just look at the characteristics of our Scores business and its profitability and put a reasonable multiple on it, you get to $17 billion, which means that our software business is not highly valued. It's just not. And so the potential for us is to demonstrate to the world that we have this amazing software business. And I have every confidence the software business is going to be worth $5 billion $10 billion in the next couple of years, next few years. So that's the potential.

Okay. And then how about like the investment priority? Do you go -- we're talking about this yesterday, down market in financial services, horizontal to other verticals, geographic expansion? Like how do you get those?

It's a great question. So we're actually quite a small company. We have about 150 quota-carrying reps that do direct sales to financial services and even that, it's really the top 200 or 300 financial institutions around the world. And then beyond that, we do it with partners. We do it with processors. We do it with other partners that give us geographic reach that we wouldn't otherwise have. And we don't think that we'll ever have kind of direct sales that can capture all the value of the IP. We're really an IP shop. We have a ton of IP and our distribute is super anemic, okay? That's the situation. And so the focus is how do we most efficiently get this IP out to the world. I think this is very interesting question. Do we go down market, Tier 2, Tier 3 in financial services, focusing on risk or do we go across to other verticals, focusing on digital transformation. Because remember, if we -- we've had tremendous success with the top financial institutions. We have pretty good penetration with the top 200. And if you can crack the code on digital transformation for banks, so much easier for the other verticals, right? Because banks have the highest needs in terms of security and fail over and disaster recovery and everything else, and heavily regulated. So we think we have something that we can offer to other verticals that we're not in today, everything from retail to telco to other verticals. So I think we'll go there. But I think we're not going to ignore going down market either. So I think that you'll see us pushing in both directions.

Do you feel like the reset that a lot of the high flying fintechs have experienced takes some of the heat for innovation and digital transformation off of the established players?

I don't think so. No. I think that the heat is very real. I mean, I think that COVID woke up all the big banks. And yes, they were feeling the pressure in 2016, 2017, 2018 from the fintechs. We're doing a better job with the app, doing a better job with how quickly you can get a loan or an application process or whatever. But I think COVID really woke them up and everyone knew the branches were shrinking, we're closing branches. That's not the future. But I think that the financial institutions recognized. They knew as a strategic priority, but I don't think they really acted on it. And I think with COVID, they just -- they all decided to retool their businesses. They have to retool them. And so the pressure is from the fintechs. It's also from COVID. It's just consumer behavior. Consumers have an expectation that I mean what could be more digital than moving money. Consumers have an expectation that their relationship with their financial institutions should be a digital one. Maybe you augment it with a branch, maybe augment it with call center, but it's got to be a digital relationship. And so that's really the pressure. That's what's happening.

And then at the same time, generally speaking, there's pressure on IT budgets, so how to reconcile the pressure to innovate with the reality of...

That's a super interesting question for us. So 15 years ago, in the last big downturn, IT budget shrank a lot. And FICO got hurt. I mean FICO's revenue went down, our profit went way down, we laid off a lot of people. I mean, it's really tough times. This is before my time, but it was tough times. And there were questions in the Board room about, should we be as concentrated in financial services as we are. I mean these -- when these downturns happen, we get killed. And there's always been a lot of talk in the boardroom about moving to other verticals and diversifying the risk. But what's happened with our offering is it is so strategic that even in a world today where IT budgets are somewhat diminished. We don't have any trouble. It's a C-suite discussion, it's strategic priority for the bank, and they may be putting other things off, but they're not putting off our stuff.

Okay. You talked about all the upside that's locked in there that's potential. Scores and software belong together?

Good question. So there are 2 ends of a continuum, right? If you're thinking about how are you going to evaluate the risk? What's one of the really big decisions that has to get made, if I lend you money, are you going to pay me back? That's the big decision. A lot of value in getting that decision right. That's why early on, even 50 and 60 years ago, we're focused on, let's make that a data-driven decision and not just be the intuition of some old bankers sitting in the corner office. So we've been trying to bring data to this decision for a very long time. The one end of the continuum is Scores, where you have a single data set with very high [ caloric ] value. I mean, what better data set to look at, are you going to pay me back in the future? Well, have you paid me back in the past? Good thing to know. That's Scores. Software says, our software business says, we're going to make the same kind of a decision about, are you going to pay me back in the future as you have in the past. But we're going to try to get some incremental value out of incremental data sets. And so you hit the mentioning returns pretty quickly. But the next data set and the next data set, and the next day set, they all provide a little bit more predictive value. And as you go to chase those incremental data sets, you get to much smaller populations because the data sets aren't available for big populations. So at the one end, we got the broad population, super-efficient credit bureau file-based scores. At the other end, you have the software business where any, and all data will make the smartest best decision possible, leveraging everything that could possibly be known about this individual. So at kind of at a theoretical level, these businesses are 2 sides of the same idea. But at a practical level, they could be separated. I mean the truth is that the investors who are interested in the scores and the info services side of things are not the same as the investors who are interested in the software potential. And I think in the fullness of time, these 2 businesses are likely to be separated, just to unlock the value. There will -- there's incremental value to be unlocked by separating them. They don't have to be together. There are some synergies, but they don't have to be together...

But for now one funds the other.

Yes. Exactly.

Okay. We've already gone through most of the time. So does anybody have a question you want to ask, Will. Mic here.

We've got mic coming.

You probably get asked this all day, every day. But -- and I think I know your answer, but just want to hear if there's any update, the AI threat to the business to credit scoring. I know people bringing up from time to time that eventually big tech will have more info about us than anyone else, so that's a threat to Scores. Could you just...

We don't see it as a threat at all. And here's why. If you're interested in whether someone is going to pay you back in the future. That's the question, right? It's a question all the lenders have. Am I going to get repaid. The first thing you're going to do is you're going to go to the data set that has the most caloric value, which is the credit card data in the credit bureau file, and it just says, you paid me in the past, I bet you'll pay me in the future. So it doesn't matter how fancy you make the analytics afterwards. It doesn't matter how many incremental data sets you bring to the party, up to and including an infinite amount, okay. Let's say, we just got in to eat there and munch, munch, munch, much, all the data available, it doesn't matter. You're still going to want to leverage that data that has the most caloric value, and that's the score. And so I see a future for Scores where it's the cornerstone of this decision for a very, very long time. Even if you throw AI on there, it doesn't matter.

Okay. I think we're out of time. So thank you, everybody. Will, maybe, you can hang around for a few minutes for folks who can't get...

Yes. Sounds good. Thank you.

Yes. Thank you.