Dan Dolev: Thank you.
Operator: Thank you. Our next question comes from the line of Bryan Keane with Deutsche Bank. Please proceed with your question.
Bryan Keane: Hi. Good morning. Max, just want to ask about Affirm Card. I know it was down a little bit sequentially, but it sounds like that was mostly seasonality. And just as we think about the longer picture trajectory of the card and the adoption of the card, how do you think about volume growth on Affirm Card? Just trying to get a sense of what the trajectory will look like given you probably have a pretty good idea what the pipeline looks like.
Max Levchin: It is entirely seasonal. That’s certainly the correct assumption. I think we talked about it last quarter, so should not be in any way surprised. Very happy with the clip. We’re still not promoting it nearly as aggressively as we feel we could because, as I just said, we have a bunch more to do just on the user interface and product improvements. Not sure I will offer a specific shape of the curve for the card growth just now, given that it’s still in this hyper growth stage. So easier to be wrong than right. But we have a huge number of things to ship there. We’re still experimenting very actively with various kinds of rewards. Obviously, you want to be careful with margins, but it’s a product that we’re very excited about.
We feel that it’s just has lots and lots of room to grow. I think one sort of a good measure like we really started to just offer it widely just if you wanted to stretch it two years ago, if you wanted to sort of take it a little bit more narrowly a year ago, it’s now very squarely trending towards multi-billion dollar business. So just from a pure like — where have you been, where are you going, not sure there are any other financial services companies I can stick a finger at that launch a card product for the first time and end up in multi-billion dollars within 12 to 18 months. And so we have a lot of — a lot of features to roll out. And I promised myself I’ll never preannounce things on these calls, so I won’t, but we’ve got some really cool stuff coming.
Michael Linford: Yes. And the only thing I’d add, Bryan, is if you look at spend on a trailing 12-month basis to capture or to cancel out some of that seasonality, we actually saw spend per user increasing at a pretty healthy clip from Q2 to Q3. So we feel very, very good about the engagement on the product right now.
Bryan Keane: Got it. Got it. And the adjusted operating margin keeps kind of beating expectations. And I think in the fourth quarter, we’re guiding of 15% to 17%. Can you just talk about what you’re seeing on the margin? And is that a good jumping-off point as we go into next year? Or are there some other puts and takes to think about?
Michael Linford: It’s a great question. Yes, I think this year has been a real year where we’ve been able to drive a pretty meaningful operating leverage in the business. That’s a — it’s a function of us really driving strong growth in our unit economics with the revenue less transaction costs growing very quickly and are flat or even reducing fixed operating expenses. And so we would expect some of those trends to play out. However, we do still feel very optimistic about the opportunities ahead of us. So there’s lots of exciting things that we’re working on that will need resources. And so we definitely not want folks to think that we’re going to be much above our framework that we gave investors in November. We think we continue to have revenue growth rate numbers that are well-above our 20% threshold that we put out there in November and would expect us to live within the operating — adjusted operating income framework that we put out in November.
Bryan Keane: Okay. Thanks for the color.
Operator: Thank you. Our next question comes from the line of James Faucette with Morgan Stanley. Please proceed with your question.
James Faucette: Good morning. Thanks for doing this this morning. I appreciate the call out on increasing exposure given the credit performance and the yields that you’re generating right now, which seems like could create some potential for near-term delinquency increases. But I’m wondering, is there a framework of how we should be thinking about GMV versus credit right now? It seems like if we look at the 30-day plus DQs on some of your prior deals that there’s still a fair ways to go to higher watermarks. So just — how should we think about that balance of driving volume growth versus where some of these other delinquency measures and credit measures may move?
Max Levchin: Good question. Thank you, James. So what I meant in my note is that we are over-earning, as Michael would put it, which means that we have capital to invest in growth. That does not necessarily mean loosening. In fact, summer in particular is a seasonally high delinquency rate time. And so there’ll be undoubtedly seasonal fluctuations in delinquencies, more likely up than down. We’re very, very concerned with delinquencies at any given time just because that’s single most important part of the job here. But having a little extra money on-hand to invest doesn’t always mean, so let’s take a little bit more risk. What it does mean, for example, that, but also you could invest that in APR subsidies, which go towards the higher credit quality borrowers and create positive selection bias in the credit portfolio.
And so we’ll definitely reinvest the extra gains. We’ll still maintain an extremely vigilant watch over credit. As we’ve always said and we’ll continue to say and behave, we don’t use credit as a growth driver or limiter. Credit is a thing we manage entirely discretely. There’s not a conversation about, well, if we just tighten a little bit, the growth will slow down. We don’t want that. We want growth as much as we can responsibly handle, but not before we have credit results that we like and most importantly, our capital markets partners like. And so we’ll look at that and we’ll manage that entirely separately and we’ll find all the intelligent ways that we can possibly grow. And I don’t think we’re supposed to offer a framework for that, but I just want to double stress that there’s no coupling there and there should not be coupling there.
James Faucette: Great. That’s great answer there, Max. And then I wanted to ask a follow up on the Affirm Card and clearly a lot of work going into the user interface and points where you can reduce friction. In previous comments, you’ve talked about some of the things that you’re doing on the customer service side and I think have also called out even using AI assistance to help improve those customer support costs. I just want to hear from you where we’re at on that process and you kind of what are the additional things that can be done to bring down cost of support versus reducing upfront usage friction and where you feel like you’re out on that process.
Max Levchin: Great complex question to spend the next 30 minutes diving into all that stuff. That’s what I like talking about. So on the card, I think I mentioned this before, so it shouldn’t come as a surprise, but we have a whole — we think of these things in work streams. So we have a work stream called SUEX, surprise user experience. And the card is a new product, as I’ve mentioned just now, and there’s plenty of surprises that we do not want our consumers to have with the carton. So we’ve been very, very busy just polishing the rough edges. And that’s a long list. And if you use the card, you will see what we’ve done there. It’s — some of it is very apparent and some of it is a little bit more sophisticated, but just to give you like a glimpse into this, because the card sometimes runs over Visa rails sometimes runs over our own, there are situations where you have to match a transaction or somebody, for example, wants a partial refund, et cetera.
And so just a logic around transaction matching going back sometimes 30 days, et cetera is fairly complicated. And the more intelligent you are in transaction matching, the better you can serve somebody, somebody who calls your caps with you and saying, hey, this transaction, I have to cancel it, showed up not as described. I have a chargeback, whatever. If you know exactly what it is, the call will go faster and the consumer will be happier. And so that sounds like a small thing, but it’s a major cost reducer, for example. And there’s like three dozen more where they came from. On the Al side of things, because I’m always terrified of sounding like too much of a nerd, because I am, we try to make sure our Al strategy is technology that is real versus PR, which is what we’re encountering a lot in industry right now or at least pure PR.
And so we’ve been, generally speaking, fairly quiet and dismissive about it. But we’ve been investing really heavily in this idea that most — certainly, GenZ consumers really love chatting versus calling and they have no problem chatting with an Al, especially if the Al is intelligent. There’s lots and lots of really complex things. One is, obviously, everybody sort of knows hallucination is a thing in Al, and you have to be very careful. But there’s a bunch of really smart solutions that people in the industry have come up with, us including, where hallucination is not a problem, and we can very, very quickly satisfy a query certainly around a question of what do you do or what’s your policy or what do I do as a consumer, if I have an issue or a problem, et cetera.
And that’s been working really, really well. And the — it’s very early, like we expect to do a lot more there. No one has yet to lose their job to be replaced by robot at a firm. So that’s not a short-term cost saving. But in terms of our ability to scale our customer service and base as we employ Al more and more, that’s certainly going to be a saving over the next one, two, three years. We expect for consumers to always be able to reach a human. I think that’s really important. But by the time they get to a human, they want to hear someone who’s an expert, who has a really deep understanding of what’s happening with that particular account, that transaction, whatever dispute or questions they may have. And you can prework a lot of that with Al. And so that’s where we’re spending a lot of our cycles and super excited about that.
Again, we try to not talk too much about it because it’s just the cacophony of everybody being Al-powered is a little too loud right now. But it’s a really exciting tech, and we’re super thrilled to have deployed it.
James Faucette: Thanks for that.
Operator: Thank you. Our next question comes from the line of Reginald Smith with JP Morgan. Please proceed with your question.
Reginald Smith: Hey, good morning and congrats on the quarter. I had two questions. The first, and I’m not sure if you guys have disclosed this, but I was curious if there’s a way to talk about like where you’re seeing your Affirm Card or the type of customers you’re resonating within your base, so thinking about income or credit band. Is there a way to kind of talk about that? Were you seeing traction in particular in any given segment? And then I have a follow-up question. Thank you.
Max Levchin: Two questions. Obviously, this is essentially a transaction card or a transactor card for a habitual revolver. I think that’s probably the best way to describe the near perfect product market fit with the card. We have all the expectations and designs to do to address many other bands of credit. But the idea of someone saying, I don’t want to be in debt in a way that I can’t predict. I like the ability to finance some things and to not finance others very explicitly, that is the purpose of the card. That’s the marketing message. That’s the story we tell to our consumers when they say, why do I need a have the card? And so that is the user. The user says, hey, some things need to be paid over time and some things need to be paid for right now. And I don’t want to commingle the two. I’m not going to pay interest on a cup of coffee. That is the buyer. And that’s what we have today, and you can see it in all the stats.
