Riskified Ltd. (RSKD) Earnings Call Transcript & Summary

Okay. Welcome, everyone. My name is Tim Chiodo. I'm the lead payments, processors and fintech analyst here at Crédit Suisse. I want to thank everyone for joining us this afternoon at the 26th Annual Credit Suisse Technology Conference. We are very fortunate to have with us today the CEO of Riskified, Eido Gal. I want to thank Eido and also Chett for joining us and making the trip here in Arizona. So it's a pleasure hosting you both.

Thanks for having us.

Okay. Excellent. We have plenty of more specific questions, but we do want to start out with -- for some of the investors that are newer -- the basic overview and introduction to the company. And then, again, we'll get into some of the revenue and margin type topics.

Sure. So hi, everyone. By the way, thank you for joining during kind of exciting second half game time. So we'll hear if there's any goals. Look, what Riskified does or what we're really good at is looking at an e-commerce transaction and understanding if it would result in a charge back or not. That's what we've been doing from day 1 and have scaled that up, and we believe we're the best in the world of doing that. And this is something that every single e-commerce merchant needs to do, okay, because merchants are liable. If they accept bad transactions, that result in chargebacks, okay. So that's something we help them with.

Excellent. So let's expand upon that a little bit. So many in the audience are very familiar with merchant acquirers, right? Merchant acquirers often have some sort of a chargeback solution or capabilities in-house and is a part of their offering. Can you talk about how Riskified is at times either an add-on to that or a supplement to that or how you might compete with those offerings at times?

Right. So let's give an example of an enterprise merchant, but we've historically shared that. We work with Booking.com, right? The way that they would be structured is that they work with several gateways, acquirers that could be like Adyen, Stripe, dLocal, and they would route transactions based on who provides them the best auth rates and best kind of cost ratio. And because you have a multi-acquirer set up, you don't want to have a single one manage your risk. And then usually you have risk as a logical layer on top of all of these merchants, making the decisions for all gateways, what should be accepted or not for fraud reasons. And that's just usually how they're set up. Another reason could be that separate from some of these acquirers, we have deeper integrations with more data points that enable us to make more optimized decisions, right? So if I'm an acquirer, my mission in life is to get you on board and routing transactions through me as quickly as possible with as little friction and data points as needed. If I'm a fraud provider that guarantees acceptance rate and the chargeback liability, I want to have as many incremental data points to make smarter and more fine decisions. So we usually end up having about twice as many data points from an integration perspective than someone like an acquired, right? And just to give you some more kind of factual context, that means that we would know the flight route, the class of fare of ticket and the passport nationality of the person buying the ticket, right? That's something that's not required if you were processing payments. That's something that could be beneficial when you're trying to optimize fraud and risk decisions.

Excellent. Thank you for that context. Let's move a little bit into your GMV growth with existing customers, and maybe bring it to life with a few examples or an example of where you might start in a certain geography or with a certain subset of transactions and how that evolves over time as effectively wallet share gains with that customer?

Sure. So -- and again, just to iterate Riskified, we focus on enterprise clients, we focus on clients with a few billion in GMV. So these clients before Riskified, they have an existing team that manages risk for them. This team is using different third-party tools. And they're deciding more transactions to accept or not. And usually the sales process with Riskified is, we will come to a merchant and say, hey, do you have an issue maybe with expanding into APAC? Do you have an issue with first-time customers? And then, we would receive some data from them, we would run an analysis and we would come back and say, hey, if you give us your APAC portfolio, which is 20% of your business, your cost structure is 30 basis points. And when I say cost structure, I mean the cost of chargebacks, the cost of staffing and the cost of tools, that's the cost structure. And the approval rate that you're getting for that cost structure is 90%, okay? And actually what Riskified can offer you is for a lower cost structure instead of 30 basis points, 20 basis points is a higher acceptance rate, right? So that's how we would initially start working with the merchant. We would then build a trusted relationship. You would see that the business is working as anticipated as expected. And since we're already integrated, we would say, hey, you know what, why just APAC, why not move into all of LATAM and EMEA, so going from 20% in this example to 50% of the merchants volume as we build the relationship and trust over time until potentially taking over the entire volume, right? And some numbers that we've historically shared, I mean, I think as a business, we're on a run rate of about guaranteeing $120 billion or slightly above that, $120 billion in GMV, right? We've shared the last Q4 that we have $300 billion in whitespace opportunity, which we define as the GMV for merchants that are integrated into Riskified but is not being submitted to us today, right? So that's kind of our upsell opportunity within existing clients.

Okay. Excellent. Before we move into the revenue growth algorithm, we just want to hit one last somewhat introductory question. Maybe just talk about some of your core verticals, but at the same time, just hit on maybe tickets and travel, which has certainly been a highlight this year.

It has. The tickets and travel is right now approximately 30%, I think up from 20% a few quarters ago and closer to 10% in the pre-COVID environment. So that's definitely been a pretty significant growth engine for us. We've been growing just organically as the industry has rebounded, but also a lot of good new additions. And I think what we've seen is that these merchants have had fixed operating costs around managing fraud and its volume went down to 0. They still have those fixed operating costs around different head count in different parts. And Riskified in a lot of ways is completely variable, right? We charge a percent of the merchant's GMV and they saw that, especially now if they're ramping back up, it's kind of a better model. So I think that's helped us gain more traction there. We've also had a lot of geographical expansion. So APAC is kind of the fastest-growing region for us, have seen good growth in APAC. Some newer categories for us are around kind of food delivery, groceries have been good for us, remittance. And that's on top of some of the more traditional kind of fashion, general retail, electronics.

Okay. Excellent. And maybe somewhat related to that, and that was a really good point that I think you clarified for folks around the fixed cost versus the variable cost structure, particularly during the downturn, which seems like it gave you the opportunity to add or at least gain share with a lot of these clients. Let's talk about the gross margins of the products and how that matters in terms of how valuable your solution could be for a low gross margin product relative to a high gross margin product.

All right. So if you have a high gross margin product, we would recommend and businesses should accept more risk to drive more incremental sales, okay? So if I'm selling luxury fashion, which has very high gross margins, I should be willing to accept more chargeback risk, more fraud in order to drive more top line units, right, because the incremental value of every unit is so high from a gross margin perspective. And we see that a lot with our fashion or different categories that have a high margin profile. When you have a fixed or lower margin profile, let's say, an OTA, I know they can't make a ton more money, they're somewhat fixed. So they need to control costs and everything up to that degree. How does that manifest with Riskified? Let me make it clear, an OTA can come and say, listen, I can pay 15 basis points, okay? That's what I can. Drive whatever amount of sales you can for 15 basis points. That's what I want you to do. That's the right thing for the business. Another merchant can say, listen, I don't mind paying 70 basis points, okay, as long as you can drive even a bit more incremental revenue for them. So obviously the more basis points we get, the more higher approval rates we can provide our clients, right? There's always that type of curve that says, hey, for this chargeback rate, what approval rate can you provide?

Perfect. This is really bringing it to light. I really appreciate that. Let's talk a little bit about this year's growth. And then I want to talk a little bit about what this means in terms of your future algo. But for now, let's just touch on this year. So the GMV growth this year has been strong in the 20% range. Tickets and travel, as we mentioned, has been a massive growth area. It's been offset a little bit by the underlying kind of other category e-commerce market weakness, and then also something that is more of a onetime in nature industry change with PSD2. Why don't we hit each of those 3 briefly, but with an emphasis on the PSD2 impact, noting that most of that is in the past, but I think it's good to clarify for investors.

Sure. So let's clarify what PSD2 is. It's a directive in the EU that forces merchants to deploy strong customer authentication, predominantly that's 3D Secure. And once a transaction goes through 3D Secure, the liability is no longer with the merchant, it sits with the card issuing bank. That happened in second half of 2021. So what that basically means is that for our business, our merchants that we work with in EU, a large part of their volume, they no longer have the chargeback liability, so they no longer needed that portion from us. So our revenues went down from them. But like all other types of churn, once you have a 12-month period, you cycle out of it, it doesn't impact the year-over-year growth rates. So that's where we are with PSD2. We've kind of shared that. It's going to be -- it was a mid-single-digit impact during 2022, and it's a nonmaterial impact in 2023. Just from a regulatory perspective, it's not something that we see on the horizon in any other geography. To give you some context, today, in the U.S., any merchant that wants to can employ something like 3D Secure and have a liability shift. They proactively decide not to do that because it impacts conversion, it's a negative customer experience, and they would rather work with a vendor like Riskified and have frictionless experience with no fraud and chargeback liability, right? So -- sorry, that was too much on PSD2, trying to move away from talking about it. It always brings up the motion. So aside from that, you're right, this year, we had strong organic growth from tickets and travel. That was offset by negative growth on kind of general retail in the post-COVID environment. Historically, we've seen about a 10% organic same-store sales, which we anticipate e-commerce would revert to that at some point, obviously recession, different headwinds, hard to say exactly when they would kind of abate and turn around. And the third part of that algorithm that you mentioned that, hey, in Q3, we grew 20%, 24% net of FX, that was because of new and upsell to existing clients. We're obviously extremely happy with that performance. It's a bit higher than we had in previous quarters. Yes, it was just a great performing quarter. It was somewhat just Q3 is a stronger tickets and travel. And because of our book of business there, it's even larger. So it's not that we're saying, hey, this is the new normal, anticipate that every quarter, but very happy with that.

Excellent. Okay. Before we move on to the topic of gross margins, I want to just follow up on -- not on PSD2 specifically, but some of the other opportunities. So you mentioned that some of those clients no longer had a need for the service due to this onetime now largely lapped regulatory change. Could you talk about what else you could still work with those clients? In other words, there are still other products that you offer that you could still sell through to those same clients where you have a relationship. Maybe just touch on a few of those.

Yes. Thank you for mentioning that. I would say that the #1 product that I'm most excited about is Policy Protect, okay? So we talked a lot about chargebacks, dealing someone's credit card and doing some of that stuff with that. Just working and talking to our merchants, we've understood that all of them have other areas where they generate a lot of losses that we can help with, and Policy Protect is, they get a bunch of refund requests and some of them are fraudulent, okay? You can call a retailer today and say, hey, I never received my package even though you received your package, they would send you a new pair or refund you, okay? And there's a lot of loss there. It's the same underlying -- and by the way, think about it, that's much easier to do than to actually steal someone's credit card or purchase the financial information online. Everyone should attempt this and see if you're successful. And if you are, send us the name of the retailer and kind of the example, it would help us as well. So that's like one use case we can help with. Other use cases around policy are around item limits, which the past week we've kind of taken Taylor Swift was a big thing as well. But it's not just tickets and events, it's also sneakers and other limited release edition. So we work with merchants on that, not let a measure on that, to decide what products are available to sell to what customers and how do we verify that it's a real customer and not a bond making some of these purchases and creating these accounts. Another example could be -- and this is a bit less about fraud, but more around merchant policies to optimize their cost structure. Do you really want to offer free shipping and returns to everyone? That's what most people are doing today. That doesn't make sense, okay? You have a few percentage of clients that have a 99% return rate on 100 transactions, you should not be offering them free returns, right? So there's a lot of logical decisions that we can make that impact the merchant's portfolio. That's [ live ] most of these use cases with a handful of merchants, and we anticipate more revenue growth from this in '23 already.

Okay. Excellent. And glad we touched on that. Let's move to gross margins. We'll get to the long term in a minute here, but let's just recap. In Q3, the gross margins were strong. They were above our estimates, they were above the Street. Clearly there's some underlying goodness from your technology, maybe there's some mix shift. Maybe just talk about what drove that really strong gross margin result?

Or machine learning? Okay. That's like -- to simplify it, it was better performance on the machine learning side to give the longer answer, which is going to end up being the same. What we see is that we consistently improve cohorts on a year-over-year basis, right? And that makes sense to you, like it's machine learning, we get smarter, we get more data, we see more chargebacks. It stands to reason that it should improve. Specifically, in this Q3, we need to go back to the previous Q3, right? And what happened is that we had a big mix shift. And when I say mix shift, it means we have the same revenue, but the merchants making up that revenue was completely different, and this is again a COVID impact because there was no travel and then there was a huge amount of travel. So that meant that the comparison was not the same, different merchants, different categories have slightly different margins with us because of the risk component. So that created an imbalance in the previous Q3. And we probably gave a more moderate target around margins because we knew we had travel, which was historically a more challenging category for us relative to general retail, but we were able to outperform expectations internally as well.

Okay. Excellent. That was a great recap for the dynamics in Q3. Let's talk about the long-term gross margins and where they could head and why that might be both in terms of, to your point, the machine learning continued improvement, but also other potential mix shifts in terms of new geographies or verticals, et cetera?

Right. So the way that we think about it internally and we would recommend investors look at it as well is, again, consistent year-over-year improvement in existing cohorts somewhat offset by new cohorts in new regions and new geographies, which could start at a higher base, right? Now obviously that goes into, well, exactly is that 15%, 20%, 25%? What mix is new? Is that mix coming from new geographies, new industries or existing verticals? How should I model that? Those things are not [ cozier ], but it's hard to give a direct answer to that right now, but we can continue to anticipate seeing that year-over-year improvement in all of the existing, which historically has created overall improvements in the business over time.

Okay. Excellent. Let's move on to operating expenses. So this has been another highlight for this year in terms of really being disciplined on OpEx, and I think it's been well received by investors. Can you talk a little bit about what have you done? What did you change relative to the way you were thinking about the OpEx base on January 1 of last year? And then, as a follow-up, as we think about the absolute dollar cost base in Q4, sort of a jumping off point into next year, is that a fair way to think about it? And maybe touch on some of the seasonality?

Yes. Happy to do it. So look, if you think about the history of Riskified, we've actually -- we've been profitable before. We've burned a very relatively small amount of capital to go public and reach where we are today. And I say that -- the reason I say that is to just exemplify that the unit economics and the base here makes sense, right, which I think it was like a big concern with the cohort of IPOs going public and some of the different things happening there. But for us, really in probably '21 and to some degree, 2021, we kind of said, hey, we have this large opportunity. We have both on the product and the go-to-market side. We want to expand our go-to-market teams to APAC to LATAM, some other regions. And we want to build additional services like policy, like others because we think there's a massive opportunity and long-term opportunity. Earlier this year, we kind of said, hey, you know what, we feel that we've sufficiently built out the global go-to-market team, and we've really grown like almost doubled head count in 2 years, which is massive. So we have a lot of R&D capabilities to build some of these newer generation products. And what we want to focus on right now, consistent with kind of what the market expectation is, is to show as high as possible operating leverage and fixed expenses to where they are and move aggressively towards profitability over the next few quarters once we reach that profitability point or clear line of sight to that or show that profitability. Just start to reinvest some of that a bit back into the business, a bit towards the operating margin. So I think that's what investors should expect us to do, [ get fixed ] cost until we reach profitability, and at that point, continue to invest some in the business, some in driving operating margins. We don't have that exact formula 50%, 30%, but we'll share that once we're closer there. And so once we made that decision at the start of the year, obviously together with just scrubbing different line items and expenses and optimizing like all companies are doing now, everything from -- that was an almost goal. So optimizing everything from cloud cost, personnel to drive that kind of fixed expenses. And I think that you should imagine Q4 as a good run rate for 2023. Q4, we usually -- there's some step-up from Q3 just related to general expenses because of the holiday season. So some different missing that, right? So if you imagine expenses to be at around 47%, 48%, consistent with what chemistry is modeling, that's a good run rate for all of 2023.

Excellent. Thank you for that answer. Related -- you alluded to this a little bit in your answer in terms of something you mentioned on the call. And you were clear to say this wasn't a firm guidance, but you said that there could be periods of profitability as soon as the second half of next year. So the question that investors have on that is, and you mentioned in terms of the improvement that might be required. But what does the e-commerce macro environment have to look like for that to come to fruition? Does e-commerce have to get a whole lot better? Does it have to be stable? Could it be a little bit weaker? How are you thinking about that?

I think one way to answer that is, just look at consensus for Q4, which basically has an adjusted loss, I think, around $8 million, off by $1 million, and just assume kind of steady margins, what type of growth do you need to see from us in order to have that negative [ 8% ] reach like profitability stage. I think you'll get to like low mid-teen type of growth rates, right? So I think from there, breaking that down, you saw on what type of new and upsell we've had in this Q3, obviously, you don't even need to maintain that. If there's some improvement in the macro, obviously, that becomes much easier. How much of that is a function of just Q4? Is it possible to target back half of the year? So those are all kind of questions under still in play. But we wanted to give investors a sense.

I think that's fair. I think a follow-up question would be for you to get to, say, mid-teens volume growth. Could you do that in an X growth e-commerce environment due to the share gains that you're having with existing customers and other aspects of your growth algorithm essentially? It's not all dependent on end market support.

Yes, correct. I think this year, we've shown that type of growth independent of e-com growth is possible. Would obviously be nicer if there was also some e-com growth.

Yes, that helps. Okay. All right. Great. I think that's a good segue into the next question, which is exactly that, which is upselling and you've had some great success with that. Maybe just talk a little bit about the types of customers that you're seeing most of that traction with you mentioned sort of the $3 billion plus. And then also, is that more volumes? Is it more work? Is it more -- sorry, is it more regions? Is it more products?

Yes. We've had the most success. Yes, we target large strategic enterprises. And this is overall a better demand environment for us than during peak COVID. So during peak COVID, all e-commerce companies were going to become the next Amazon and the cost of capital was 0, and it was fine to build a large internal team to do everything. And you were growing 50% anyway, so an additional 3%, 4% was not that like important to you. In an [ intravenous ] environment where there is a much larger focus on justifying expenses and understanding what's core competency versus what could be accomplished better by partners for a better cost structure with better results, we're seeing more demand for us, unlike other -- maybe some other IT services where it's like an incremental spend or actually replacing an existing spend for a better cost ratio. So we haven't seen any type of like lengthening of sales cycle or anything like that, right? Because you're basically saying, hey, it cost you $10 before. I'm going to do the same job for $8 or something to simplify that. So the better demand environment has definitely helped us. It's usually to your question about what we upsell, is it more segments or more products? This year, it's probably been more volume, more than anything. So going from 30% volume share towards like something like 100% volume share is the most significant increase. What else? Did I miss anything on that?

Volume would be probably covered by moving into new geographies.

Correct.

I think that's fair. Okay. We don't have a bunch of time left, but I have one more that I really want to make sure we work in, but I also want to open it up just to the audience. I think we have time for probably one from the audience, and then I'll wrap up with a final one. Does anyone want to chime in? Please go ahead.

Could you just share your long-term...?

Microphone coming.

Could you just share your long-term vision for the company?

Sure. Happy to. The way I think about it is in terms of GMV and take rate. And I think we're very happy that we're above $120 billion in GMV. But when you think about e-commerce, that's still like a multitude of trillions opportunity. So putting e-com growth aside, there's still massive market share gains. And if we go back to how we started this presentation, we're the best in the world at looking at every single e-commerce transaction and deciding if it's fraudulent or not across industries, across geographies. That's what we want to be doing for trillions of dollars of GMV. And we think there is network effects here. And the more we do it, the better we can do it, the more resources we have, and we're doing it in an increasingly more accurate way than anyone else in the market. So that's part 1 of the strategy. Part 2 is, together with using that engine to decide if a transaction is going to result in a chargeback or not, what other common pain points do e-commerce merchants have -- enterprise e-commerce merchants that we can help solve for them, how can we generate more value using our existing data network effect. And an initial use case could be the policy examples that I shared, okay? So I think that's kind of the overall strategy.

Excellent. I'm going to try and wrap up with this last one. So on the last call, you talked about a big new customer that you won. So it was described as one of the world's largest secondary ticket marketplaces. It was also a competitive RFP process. So maybe just talk a little bit about that. How many competitors are coming into these RFPs? And what are you submitting as a part of the RFP process that is really helping you stand out? And what are the others not submitting?

Yes. No, that's a great question. It's usually around 2 competitors. And the most important part of these POCs is...

2 others or 2 total?

2 others.

Okay, so 3.

So the most important piece is the POC and the proof of concept, and that's why we predominantly win, okay, because of the accuracy of the solution. The reason we love the enterprise is that they don't look for a bundle play that one vendor is giving them. They choose deep, best-of-breed solutions, and they test them and they go with the most accurate one, and then you can also price accordingly. So when they run these pilots, they -- sorry, again, it's not technical. They give you a test set of data for you to train your models on with decisions, you train your models. Then they give you a test at with no decisions until you give me back the decisions they already know. And then they can just do a comparison between merchants -- sorry, between vendors and say, okay, this is the most accurate one, all right? This is the least accurate one. This is in the middle. And now it's more of the financial question. This one is so much more accurate, but they're asking for a bit more money. Do I want to give it to them? And that's predominantly how we win these POCs.

I'm really glad that we were able to squeeze that question in, because I think it's important because the RFP process is maybe a little bit different than maybe other vendor or partner type of RFP processes because it literally is a meritocracy that is based on data results and there is a trial in any cases, maybe not all, but it sounds like that is something that is somewhat common.

Yes, you're right. I mean, it makes sense to me, but I don't have that perspective of other companies, then maybe it's done more about who writes the nicest answers, I guess.

Excellent. We do have maybe time to squeeze in a quick one. If anyone wants to hop in?

This is probably taking up a big like step back, but just any thoughts about where kind of e-com is holistically as we're looking forward for 2023 and what your overall market looks like?

I think, again, I would echo that. I'm sure that there is some impact of recession in consumer spending and inflation. I think that impact is going to be different across geographies and industries, right? So like when you're trying to build that out into your model, it's going to become very complex very quickly, probably more pronounced in Europe as pronounced in some other categories like luxury fashion. And when exactly it ends and we revert to more normalized times as a backup of next year is the beginning of the following year. I think that part is harder for anyone to understand right now. But I think long-term, that's probably what we'll revert back to, and we'll probably see more industries like 2 years ago, we're not part of e-commerce being a more significant part of that, groceries, food delivery stuff, things of that nature.

Excellent. Eido, Chett, thank you both for making the trip to join us here. On behalf of everyone at Credit Suisse, we really appreciate you being a part of our conference.

Thanks a lot.