Figure Co-Founder: How Figure Became A $10B Business | Mike Cagney

Thanks for having me.

I've been doing great spending time in New York enjoying the weather.

You know what, I like to look at the world as a half glass full.

Yes.

Yeah, look, I think blockchain is a massively disruptive technology, maybe the most disruptive technology we'll ever have in our lifetimes, more than the internet, more than the spreadsheet, more than things that, you know, we generally attribute to significant disruption. And the idea of displacing trust with truth, you know, through native digital assets and what that affords in terms of both the way those assets are traded bilaterally, self-settle and the way those assets can be financed in terms of, you know, encumbrance to the asset and, you know, having the asset-based lending construct through DeFi to me are game changers. And so the potential for the industry is absolutely massive. And it's interesting because when I see all these people doing things at the periphery in blockchain, right, like Visa or DTCC or these like the reality is blockchain disintermediates their entire business. And so part of the challenge that we've had with blockchain is you know normally incumbents disrupt at the margin but blockchain is so disruptive there is no margin. It's a complete wholesale rebuild. And then, you know, certainly we had four years of regulatory morass that, you know, basically killed a lot of development that we're now making up for and I think starting to accelerate in terms of what's happening. But there's still some necessary conditions for blockchain to be successful. And the most important is it's tiny, right? And when we look at it, you know, we think about, you know, a DeFi protocol with 12 billion dollars of stable coin on it, we're like, 'Oh, that's massive.' Well, that's completely irrelevant in the scheme of wholesale capital markets, right? Or when people are like, 'Oh, you know, we've tokenized a billion dollars of stock.' Well, it's a $133 trillion market cap industry. And so, you know, we're still just this tiny, you know, almost irrelevant pool that to make relevance, we're going to have to bring a lot of those tradfi dollars over.

Yeah, so again I think going back to two things: to transactions and to lending. So on the transactional side every marketplace is massively intermediated, right? So when we trade stock there's seven parties that sit in between a buyer and seller. When we swipe a debit card there's five parties that sit in between the buyer and seller on the interchange network. Blockchain has the ability to distill all those marketplace transactions down to bilateral, just buyer and seller. And so by disintermediating all those people to sit in the middle, you're freeing up trillions of dollars of what's just renting market cap. And that accrues to the benefit of the individuals transacting, but it also accrues to the benefit of the blockchain, to the ecosystem that's doing that disruption. On the other side of the equation on the lending side, you know, you had mentioned I'd started SoFi or co-founded SoFi. Part of why I was doing that was I believed in this notion of connecting sources and uses of capital, right? So the original thesis of SoFi was we get alumni to invest in student loans and, you know, there was a natural reason why they'd have that affinity and all these externalities would come out of it. And, you know, we were sort of a victim of our own success because the loan was so popular it just massively outstripped the pace at which we could acquire our alumni capital. So we went and we got warehouses and, you know, started securitizing and doing all the normal wholesale capital market stuff. But what blockchain is giving us a chance to do is a second shot at this. And DeFi, the idea of being able to put up collateral where, you know, for certain it is the asset. You can get true perfection. You can get a UCC lien on the asset, right? There's a whole legal construct that's supportive of this and be able to liquidate that asset. It changes the lending construct from I'm underwriting the credit to I'm underwriting the volatility and liquidity of the asset and that massively opens up how financing markets work. And we've seen this in a microcosm with crypto financing. You know, if we look at a platform like Figure for example and you're seeing it with Figure with democratized prime where we do it with loan participation, right? And, you know, starting with HELOC, now we have auto, now we have small business and soon we'll bring on receivables and other categories. And I think that's going to change how finance works in terms of the financing side of the capital markets.

I think Figure has been a great illustration or demonstration of what blockchain does. So, you know, there are people that are critical of Figure always say well Figure is just a lender and like well show me a lender that operates at 50% EBIT margins, right? And, you know, we do that because of massive efficiency we get from the technology and, you know, have been supporting uh, you know, there's a sort of rule of 40 for SaaS companies of, you know, combination of EBITDA and growth we're at 140, right? So and, you know, and we're still in what I consider greenfield markets and, you know, massive upside of growing off of that. And so, you know, if you're sitting at a bank or, you know, at a traditional intermediary, the reason you should care about this is, you know, there's now a very clear demonstrated proof point as to why this technology is going to matter and how it's going to change things. And, you know, you could decide to lean in early and be a part of that or you can be reactive to it. But it's coming and, you know, and what's been interesting is now that blockchain is legal, the big banks are leaning in. You know, I'm seeing actually more innovation coming out of Goldman Sachs than I am out of the old crypto industry. Um because, you know, now they can lean in and do it and they realize they can make more money. And crypto itself is an important construct here because I'm talking a lot about blockchain, but decentralized blockchain doesn't work without a cryptocurrency, right? You need something for that network. If it's proof of stake or proof of history, you need something to be able to stake up as value across the validator ecosystem. And so a lot of people don't understand that tie-in or that association, but, you know, I'm talking about the value proposition of blockchain, but that decentralization, which is core to that value proposition only exists if you have a native crypto asset to be able to support the decentralization of the network. And so that's why you should care about the whole ecosystem.

Yeah, it did. And again, it kind of goes all the way back to this idea of connecting sources and uses of capital, right? And as I said, SoFi took a trajectory that was a very successful one, but was very different than the original intent of what we had. And what we were able to do with Figure is kind of reestablish that construct, right? And reestablish this idea of DeFi lending. Now, we couldn't get there right away. And, you know, in the beginning, like one of the interesting things people don't realize about Figure is in the beginning, I had to leave money on the table with my loans because the hedge funds were like, 'If we don't have to deal with blockchain, we'll pay you more.' And I was like, 'Well, the only reason I'm doing loans is so...'

So, I'm origing loans on blockchain and the hedge funds all have to do that, the original buyers are all having to deal with the blockchain and they all hated it and, you know, hated the concept of it, hated the idea of a wallet, like all this stuff.

No. And we would and we because we were using the loans as a forcing function to get them to use blockchain and they would say, look, I'm paying you 102 for these loans. I'll pay you 103 if I don't have to deal with this blockchain. And my capital markets people are like, ah, you know, let's just forget about the blockchain. Just make it Figure 2 or, you know, SoFi 2.0. And I'm like, no, that's not why we're doing this, you know? We're doing this because there's an enormous opportunity and, you know, in the beginning it's hard and there's a lot of friction but, you know, ultimately that will invert and we'll get that efficiency. And, you know, on the capital market side we went through the same thing with the DeFi side, right? In the beginning what it cost us to finance on DeFi was actually higher than what it would be, it was mixed in certain pools, it was higher than what it would be in the wholesale capital markets. That's now come down to a point where that's no longer the case. So, you know, you end up having to leave dollars on the table in the beginning to drive adoption, but then you hit that inflection point and that's where the real efficiency and the moat happens.

Yeah. So and we could spend a whole day on the dynamics and comparison of the two. Bring out the whiteboard. Exactly. We could. So if you think about the traditional process of originating a loan, you originate the loan and then you put it on a spreadsheet. You send it over to your bank to finance it in a warehouse. And the bank will let you send that spreadsheet over once every three to five days because you give them a spreadsheet full of loans. They give you a bunch of money and then they spend three to five days figuring out if you lied to them about any of the loans. And this is why you've had some very high-profile fraud cases like First Brands where they sent two spreadsheets out to banks and got twice as much money and, you know, that's part of the inherent problem. So how does that compare to blockchain? Well, blockchain the loan is on the chain and there's a digital perfection to the loan through something we have called DART. It stands for Digital Asset Registry Technologies but it gives you either a UCC 8 or UCC 12 digital perfection of the asset real time and that is the representation of the asset there isn't anything else that has that asset so there is no double pledging, there is no ability for, you know, me to sell a loan that I'm borrowing against or, you know, what have you. So that dramatically reduces the amount of risk the warehouse lender takes and it increases the frequency in which you can pledge those assets into the warehouse that reduces the amount of equity capital you have to hold on your own balance sheet because, you know, I'm not having to finance three to five days worth of loans before I send over a new spreadsheet and so there's big efficiencies that come off of that reduces your cash needs. So that's sort of the aggregation point, right? Then on the transactional side, um, you know, in the construct of whole loan sales for example there is no liquid capital market for whole loan sales historically other than Fannie and Freddie mortgages because Fannie and Freddie always have a bid and that guaranteed liquidity allows for a liquid marketplace for those loans to trade. But, you know, a jumbo mortgage or a student loan or a personal loan. There's no liquid capital market for that. And what happens is those loans all trade bespoke. So, you know, I might do a deal with Apollo to sell them $500 million worth of loans. They're going to ask for particular reps and steps in that contract and that's different than the ones KKR is going to ask for. And the resulting problem is I now have these two pools of loans that are not the same and they're not fungible or homogeneous, right? So there is no liquidity scale that happens. They're just one-off illiquid pools of assets that come out of this origination process. So what we did is we said, well, look, you know, we have over 380 partners that originate assets on the Figure ecosystem for mortgages. They all use the exact same technology. So it doesn't matter if it's Loan Depot or Guaranteed Rate or Figure. Everyone's originating the asset the same way. And so that homogeneity is a necessary but not sufficient condition to create a liquid marketplace. We have a capital market called Connect. And in Connect there's one contract that everyone agrees to for the buying and selling of loans. So there's no one-off negotiation.

It happens all the time. And, you know, at one point a massive insurance company came to us and offered to pay us two points on all our loans and we would underwrite to their box.

And we said, 'No, we're not going to do that.' And...

Very sad. And so because...

We lose the homogeneity, right? And so what we have is we have homogeneous assets and a common agreement to transact those assets. And then the last piece of the equation is we feed all the remittance of the assets into the pool real time. So I know for certain whether a loan's delinquent, whether it's performing because, you know, typically you trade a pool of loans, you spend two weeks of diligence figuring out well are they what was represented to me. Well blockchain gives me truth over trust, right? So I know real time what the truth is in that transaction. And so the combination of those three factors supports liquidity. And so we run a marketplace Connect that as far as I see is the only liquid private credit capital market that exists outside of Fannie and Freddie mortgages which are quasi private, right? And so that liquidity is very important for us because it brings down the spreads at which the loans trade because they're liquid, not illiquid. It actually allows us to compete against Fannie and Freddie on a pricing basis which, you know, historically has been just an unassailable process to try to do. And so it's kind of a critical step for us that liquidity is very important. And then the third step on securitization, you know, securitization you think about all the cost that goes into securitization. So, you know, the amount of audit and QC you have to do and third party review of loans well blockchain because of the immutability of the technology eliminates a lot of the third party review that you have to do to get your pools rated. Um so there's efficiencies that come out of that. If you look at things like the calculation agent, you know, that actually has to go through and calculate the cash flows and distribute, you know, payments to the bond holders. Well, that's a smart contract, right? You can do all that through technology. But most importantly, you know, what happened in 2008 was the market went bad and normally what you do with a mortgage securitization is when the loan stopped performing, you go and you work out the loans, you provide some payment plan or some reconciliation with the borrowers to keep them current. Um, but, you know, when a modified loan and, you know, that's a standard servicing process that one does and it's actually critical during periods of stress. But once the loans go into a securitization, you can't modify the loans without consent from the bond holders. And the problem we had in 2008 was no one knew where the bond holders were. And so had those bonds been on blockchain, we would know every wallet address and be able to push an NFT to every wallet address to do quorum of, hey, we want to do this modification. It's important. Will you let us do it? And actually drive that vote. And so, you know, there was a very high-profile regulator, um, it's actually Sheila Bair, who was looking at what we were doing. She said, 'Look, what you have wouldn't have prevented the crisis, but you would have known where all the bodies were buried, right? And we would have gotten out of it a lot faster.' And so, you know, so it has really significant downstream benefits as it relates to the securitization side.

What's already live, all you want.

So, this it's a really important point. It's really important to understand Figure. Um, so when we took Figure public, uh, you know, Goldman who was our lead left in that transaction because we took Figure public twice because we did it once on the blockchain, which was a cool thing. But when we took it public the first time, you know, Goldman was pushing us really hard to go and emphasize our loan origination system and, you know, how important it was. And I was like, look, the LOS, A, we give it away, so it doesn't even drive revenue for us.

Yeah. And B, like anyone can build any loan origination system they want. Like I can make one that if your name's Mike, you get a loan. Well, that's a zero-cost origination system. Great. But no one's going to buy that loan. And so what Figure has in the moat is the capital market, right? It's Connect. And so what we bring these 380 partners isn't a better origination system or better technology, although it is better technology, but, you know, that's not a long-term moat. You can replicate that technology. What we're bringing them that is very hard to replicate is a capital market where they can guarantee the ability to sell those loans and that's key.

Figure Connect.

Yeah. And this is the key focus point which is by having this capital because if you're an originator and let's say you're doing HELOCs for example and you're really successful you're up to 200 million a month or 300 million a month. Well the capital markets will always shut down and when they shut down you can't sell the loans but you can't turn off the funnel. So do you have a balance sheet that could take two three months of production? You know, that could be a billion dollars and absorb that. And when we were at SoFi, this happened to us all the time, right? We had a billion and a half dollars on the balance sheet and we would periodically run out of money because the capital markets would shut down. And what people who aren't in the market don't understand is you have these warehouse providers, you have these guaranteed flow by, no one will send you a wire when things are going crazy. And so even if they're committed to take your loans or committed to finance them, they'll call you up and say, 'Don't send us anything because you're not getting any money.' And I don't follow that. Yeah. So, let's say that I have a warehouse with Jeff, right? And Jeff has a billion dollar warehouse for me. And what I can do is I can pledge up to a billion dollars of loans into that warehouse. And that's great. And normal, you know, during normal times, it's great to have that liquidity. You can, you know, send as many loans over there as you need to send over there. When the capital market shut down, when, and what I mean by that is there's volatility in the market, there's concern, things stop trading or they start trading at very wide spreads, your warehouse providers, among others, will call up and say, 'Hey, I know we have a contract to do this, but don't send us any loans because you won't get a wire back.' And that's just how it works. And you can sue them. You can do whatever you want, but that's somewhat of an immaterial thing at the time because at the time you're worried about your liquidity.

100%. And the problem is that the more production you do, the more balance sheet you're going to need, the more equity capital you're going to need for when this happens, right? Because you can't turn off your origination engine or you kill your franchise. And so, you know, you're in a very tough situation. So, you know, if I said, 'Boy, I'm going to go all in on HELOC or non-QM mortgage or what have you,' that's great, but if I really want to do a billion dollars of that, I probably need a billion dollars of capital on the balance sheet, maybe $2 billion of capital. So, and that has to grow basically linearly with your production, which means you never have margin expansion, right? And so what we've done for the industry is we and the only marketplace where you do have a guaranteed takeout historically has been Fannie and Freddie, right? Because you can always sell them a mortgage. You might not like the price they pay, but they will always buy it from you, right? And so what we've done that no one else has been able to do is we've stood up an alternative liquid capital market. This was our deal with Sixth Street where Sixth Street gave us capital to operate a guarantor where that guarantor functions like a private version of Fannie and Freddie. And we can create a liquid marketplace where if you're part of that Figure ecosystem, you're one of those 387 partners that we have right now or whatever the number is, it's growing every day. Um, you have access to that capital market and that liquidity.

Yeah. Well, Figure Connect was I think you said it was 1.6 billion out of the 2.9 in.

Yeah.

Oh, there it is. We got it something like four billion this quarter from Figure Connect.

Okay.

We don't lend at all. We I mean, we have a tiny little slice, but we're not a lender. We are a marketplace.

They're getting it. And so it's kind of a slow process for us to talk to the market. And one of the challenges that investors have is there's nothing that looks like Figure. And so, you know, if you kind of look across, you're like, 'Is Figure SoFi?' Well, no, it's not because we don't have a customer, you know, a retail franchise, nor do we have any aspiration to do that. Um, is Figure Intercontinental Exchange? Well, there's aspects of our business that aspirationally are trying to compete with us. Sure. But not exactly the same thing. And so they struggle with what is the metric in which they should track and historically they've tracked take rate which, you know, and take rate is a great metric for marketplace businesses where you have consistent revenue. So like Uber and eBay are good businesses where take rate's a metric that you care about in terms of thinking of valuation for the business. For Figure, it should be contribution margin, not take rate. And the reason being is let's say that we make five points on a $50,000 HELOC, right? So the take rate's five and we make $2,500. Well, we can make three points on a $200,000 first lien HELOC and the take rate's 3% but the revenue is $6,000. The key is it costs us the same to get both those loans through the door, right? So we make a lot more money at a much higher margin on the three points on 200,000 than the five points on 50,000. So on the surface you'd say, 'Oh, the take rate fell from 5 to 3. That's horrible.' But the contribution in EBITDA margin and gross EBITDA went up. So that's actually very good.

And so we need to do a better job educating the market about what it is we're doing. And by the way, we're moving into this first lien HELOC space because a first lien mortgage space is 25 times larger than the second lien traditional HELOC space. And so, you know, our TAM is actually 25x of whatever you thought it was six months ago. Because we're now competing and actually winning business in first lien.

It's the marketplace. But you hit on another super important point which was, you know, SoFi was born out of this concept of cross-sell and the super app, right? The point of SoFi was we had these customers that we called Henrys, high earner not rich yet, right? That were super valuable customers and, you know, the idea was we're going to build a relationship with them and we will be everything for them over time, right? We're going to be their bank, we're going to be their broker, we're going to be their crypto platform and that made a lot of sense. And that still is the model that they're pursuing and that Robin Hood's pursuing and others are pursuing. But what Web 3 opens up is a different kind of model which is a wallet-centric self-custody model. And today the problem is we have two ways that people access technology. We have the super app and then we have, you know, sort of the MetaMask and the Phantom wallets where I'm attaching to DeFi protocols but I'm using the wallet to log into someone else's protocol. Right? So the experience is inconsistent as I'm going from protocol to protocol. What I think we need to do is blend these two together, which is we use a wallet, it's self-custody, but the protocol should act as true distributed applications. Like there should be an API or the equivalent of an API front end on top of those protocols and everything should drive out of your wallet. The wallet should be your cockpit. And so when I go to these different protocols, I'm not going to a different UI/UX. I'm running it out of the wallet and I'm just interfacing with them through the back end. Like, you know, how disparate or different is a DeFi construct? It's very simple, right? You know, I'm putting up collateral and borrowing against or lending against it. Like, you can easily feed that through as an API and create a consistent wallet experience. And so I think that is a necessary thing we have to do to get those SoFi customers and Robin Hood customers to come over because then it'll look like what they have, but it's a much better deal for them because they control their data. They control their assets. They control where they bank, where they broker, where they do crypto.

Oh, yeah, they do. Yeah. Because they're going to get the best deal by doing that, by being able to cut across. And by the way, they're going to have AI agents in their wallet that are going to do it for them, right? Where you're just going to say, 'Go get me the best yield where I have zero risk or minimal risk, right? Or go get me the best yield where I take this much risk, or go find me the best venue to arbitrage Bitcoin, whatever it might be.' And so, you're going to have these armies of intelligent agents that sit in these wallets that basically will take advantage of that DApp ecosystem and be able to create a huge amount of utility for the users.

So this is a skunk works project we're working on. You're going to hear more about it from us. Uh but we are working on exactly this concept including with a center AI agent which, you know, comes with a set of skills like it can produce your 1099s for you. It can do tax loss harvesting for you. It can do a bunch of things like out of the box that you don't have to train it up to do. And then there'll be things that you're going to train it to do as well.

Well, it's look, I think that we have all the ingredients for a wallet that Tradfi can embrace, right? So, we have Yields, which is a yielding stable coin as fiat on/off ramp. Your dollars always earn. We have Demo Prime, which, you know, most of my Tradfi friends want to be on Demo Prime and make six to 7% lending against Figure HELOCs unlevered. We have Hastra where, you know, you're looping out to mid-teens to 20% returns. We have Open which is the first public equity and now we have, you know, Open World's the second company to go on there but I have a list of companies that are going to come on so you'll be able to connect to that through the wallet. We have a passporting function where your KYC including your accreditation if you're accredited or what have you, it travels with you and I'm working on different entities being able to accept that. And you will have this AI agent and the AI agent will do everything from as I said produce your 1099s at the end of the year based on all the activity that you've done to buy you tickets to San Francisco.

I don't know if I call it a competitor. I don't even know if I'm going to call it a wallet.

Yeah. And as I said, I think the issue right now is and I think that's the right way for Phantom to look at it right now. What I think is there's an intersection to you said a convergence right of these things which is you want the in-app experience that Robin Hood has but you want the self-custody and ability to choose venue that Phantom affords you.

Right.

That's right. Um if your wallet holds all of these things, Figure stock and what's your the state Yields, crypto loan obligations, it's got your KYC. Can it also negotiate credit for you?

Yeah. And you'll have a card embedded into that. The card can do unsecured and secured lending. Um you're gonna, you know, because you have all those assets be able to do cross-collateralization in a way that you've never been able to do before and get credit for that. So you're going to have a very different lending ecosystem where, you know, with a HELOC you back it with your home but why not back it with your assets and, you know, have the ability to lower your cost of funds because of that. So, yeah. So, credit will and, you know, again, credit and I use the term agentic lending, which everyone likes to use because it sounds pretty sexy, but, you know, it's algorithmic lending. Um, credit will be an integral part of this.

I think this is the bank. So, if you can hold your cash in an asset that pays you interest backed by treasuries, right? That's most people would agree that's either as good or better than FDIC insurance, right? You've got a card on it, so you have easy off-ramp fiat transaction because the biggest issue we have with stable coins right now is we still can't go to Starbucks and buy coffee with them, right? So, okay, fine. Put a card embedded in there, a credit card, um, and, you know, at the end of the month, you can just roll your coin to pay that balance off and so you get the free float for the month or not, or you finance it. Um, and you have access to investment opportunities. And what's interesting and sort of the false dichotomy that comes out of the banking industry and all this is, you know, they're like, 'Oh, well, you can't pay interest on stable coins because if you do, you know, all these deposits are going to leave the bank.' Like the Treasury did a study that said $6 trillion would leave the bank system, right? And there's $18 trillion of bank deposits. So, that's a pretty big number. And just, you know, as a data point, in late '22, we had about a trillion dollars leave the banking system because of the Fed raising rates and it caused mass chaos in the capital market. So, that whole thing I talked about about capital market shutting down, late '22, early '23, they shut down, right? And because the banks all of a sudden weren't flush with deposits to fund these assets. So they start selling the assets in fire sale prices and the market's, you know, falling apart. And so but the false dichotomy here is okay we're all going to take our cash out of the bank in the form of stable coin but then we all are going to reapply that onto DeFi protocols to lend directly and we're going to take the bank away as the intermediator for capital. And by the way we might do affinity lending we might lend to, you know, our area, our geography where we are, to our school, to alumni from our work, you know, whatever it might be. But we're going to build out a very different capital allocation ecosystem that I think replaces the bank. And it's, you know, in economics there's a term called Pareto efficiency. It's where two parties agree to something where they're both better off. This is a classic Pareto example where the borrower will be better off, the lender will be better off because the bank is a rent-seeking capital intermediary.

Yeah. So with SoFi it was very brand centric and with Figure we were very conscious that we weren't going to go after retail and build a strong brand because we knew we wanted B2B to C.

Yeah. And well and it's not that I think retail is bad. Retail is just a different beast. And so, you know, we didn't want to compete with our customers, right? And so it was important for us not now, you know, we still have a small direct to consumer platform at Figure and we do that for a lot of experimentation we also do that to keep everybody in line because you're like well if you don't do it we're going to do it. Uh but, you know, generally we're not trying to do the same thing but it's one of those, you know, brand is so important and germane to the success of a retail platform and it will drive people insane because it's one of the hardest things to quantitatively measure and so, you know, your marketing people are like well I need to spend this much on brand and you're like what's the return? They're like how can you measure it? You know, it's brand. It's like existential and so there's this constant push and pull tension that happens especially between finance and capital markets people and then marketing people as it relates to this. Um but, you know, you need a...

The most important thing, and what I think most people miss around direct-to-consumer marketing, is context. You need to have as personalized of an outreach and engagement as you can. I always laugh because of all the different strategies that go through the cold emails that people have. For a while, I was getting, 'Hey, did you do this at the Stanford Graduate School of Business? Wasn't that great? Our security service, blah blah blah.' And I'm like, it's so contrived and lucky. Versus, 'Hey, this was really cool. I also participated, and if you ever have time...' Generally, marketing tries to do context, but it comes across as clumsy and contrived. But if you can do good contextual outreach, you have massive adoption.

Well, I didn't make that decision. I did the first two Super Bowl ads. Let me tell you about the Super Bowl ads and I'll tell you what I think about the stadium. The first Super Bowl ad we ran was called 'Great, Not Great.' It was like, 'Are you great enough to get a SoFi loan? Probably not.' It was kind of antagonistic, a little divisive.

Let me go through it. We had an experience about a year before where Yahoo wrote this piece on SoFi and said SoFi lends to rich people. Rather than it having a negative impact on our funnel, it was one of the best days we'd had up to that point in the company's history because people are like, 'Well, I want to be rich, and I'm good enough.' That antagonistic aspect proved to be a very effective top of funnel. We run this ad and the critics ranked the ads from 32 or whatever it was and we were number 31. Number 32 was this irritable bowel syndrome ad where this little dude's running through your colon and gets pooped out. We were one ahead of that and below everyone else. But when that ad ran, our top of funnel blew up. It unfortunately happened during a period where the capital markets were shut down. We literally ran out of money in two days. We were just lending out so much.

Yes. Sometimes you're upside down. That was a very interesting demonstration of how that works. The next year we ran a Super Bowl ad. This ad should have been outstanding because it happened during overtime. The NFL didn't price it right and basically we only paid if the ad ran and we paid less than we normally would have paid. It was just this goofy thing because everyone's watching the game. We run this ad and it's this feel-good like everyone's in this together and we're gonna fix our finances. You can't even tell we ran an ad.

Nobody cared. The funnel didn't even move. You learn a lot about how you can't go into it thinking you're satisfying critics. You're trying to drive your funnel. And two, let the data show what you do. It doesn't matter what you think or what you feel. If you have to go to cocktail parties and people tell you you had the 31st performing Super Bowl ad out of 32, so what? Because you're running it to get loans in our case.

Blocky.

It was great. People loved it. I think it was the first crypto ad I ever saw and I was like, 'Oh my god, we're mainstream.' It invoked a massive debate on Reddit because somebody who didn't understand what blockchain was was like, 'What a horrible representation of chains and a lender and you're going to be enslaved to them.' And then someone was like, 'You mean block chain?' It had this huge debate on designs that suck or something on Reddit and people were just going off on it.

Oh yeah. Yeah. People are blocking.

I think marketing is kind of interesting because with AI, your performance marketing can basically 90 plus percent of it should just be driven through AI. AI is going to do all your channel management, all your experimentation. Whether or not it does the actual copy or not, you have to think about that. I think about earned media, which is your air cover. I think about performance marketing, which is your acquisition. And I think about brand, which you could also say is air cover, but it has much more value than that. Earned media is always going to be a person-to-person activity. Your AI bot isn't going to go reach out to the journal and try to suggest a story. Brand, a lot of that content creation and management and testing can be AI-driven. And performance is going to be almost all AI.

This is the big discussion we're having about how do we do that business. Is that a Figure business? Is that a business independent of Figure that Figure is an owner of? We don't know yet how that plays out. That's one of the big concerns that we have. We have no aspiration at Figure to build a consumer brand. Yet we need to solve this problem, and I think this is an existential problem for the industry. If we cannot get these dollars over, we're going to be middling in irrelevancy for the rest of our lives.

Yeah. From TradFi. We need the SoFi customers, but we also need the hedge funds and the asset managers.