Thank you guys for showing up today. We have the pleasure of having Eido Gal from Riskified. For maybe those who do not know Riskified, can Eido, could you maybe provide just a brief overview of what Riskified does?
Sure. We started by helping merchants manage e-commerce risk, looking at incoming transactions and understanding if they're fraudulent or not. Over the years, we've expanded to also helping them manage their different store policies and find abuse. For example, if someone is claiming that they never received their package, even though they did, to manage their chargeback representment process with the banks when someone does file a claim against them, and also to protect consumer accounts.
Awesome. Maybe you talk about the specific problems, but maybe for reference, it'd be helpful to know how big this market actually is.
Sure. The problem starts that, if you're not familiar, merchants are liable for card-not-present fraud, right? Basically what that means is that if I steal your credit card, I order something from an online merchant, I receive the package, you see it on your statement, you ask around at home, "Hey, did anyone order something from whoever this merchant is?" Everyone says, "No, it wasn't us." You contact your card issuer and you say, "It's an unauthorized usage. Someone stole my credit card." They say, "Hey, no problem. You're protected," right? They pay you back that amount, and that amount is taken out of the merchant's account. Merchants are liable for card-not-present fraud. On an aggregate basis, they end up paying hundreds of billions, okay, because of this.
The sample P&L for an e-commerce merchant might look like this: 25 basis points in chargebacks related to fraud, another three basis points in staffing. Those are the people that need to manage this process. Another two basis points that they pay to tools to help them build better rules and modeling and detection to kind of identify that fraud. When we started Riskified back in 2013, we kind of said, "Hey, you know what? We do not think that we want to just be a tool that they use internally. We actually think we can do a better job of just doing an end-to-end decisioning network because we have better machine learning capabilities." Back then, it was machine learning. It was not AI yet. We have better ML, we have kind of better network effects, and we have better engineering resources to solve this problem.
We want to provide our customers guaranteed performance, right? We come to our customers and we say, "Hey, we're going to guarantee your cost." Basically, if we approve a transaction and it results in a chargeback, we would pay you back. We also guarantee the approval rate that we provide our merchants, right? Because really, the easiest way to take fraud down is to take approval rates down, which is obviously not kind of very helpful if you want long-term great relationships with your customers. We guarantee both the cost, and by doing that, we're able to replace the entire 30 basis points of cost and not just what was historically the two basis points that merchants paid their service providers. That was a long answer, but hopefully it gives more context.
That's very helpful. I guess you talked about solving the problem, but how is it that Riskified looks at this problem and effectively solves it on the behalf of merchants? I guess building off of that as well is how are you compensated? You mentioned 30 basis points. What is the take rate or how do you, I guess, look at solving for that take rate component to be compensated for the risk?
Sure. Let me start with the technology, how we do it, and then kind of the economics and how we work with our clients. The technology starts with a very deep level of data capture per transaction. We have a beacon that sits on the merchant's website. We see the different pages that you're browsing. We understand what products you're looking at. We understand if you're creating an account, things that you add to your account, things that are in your cart. We have the checkout page information. We also have merchant-specific data that they pass to us. For the OTAs that we work with, for example, they would pass us the flight route, the cabin and fare, right? That is very specific merchant-related data per industry. Afterwards, we also have data if you contact the support team. What did you say?
Did you request an address change? We have information if you've filed a chargeback kind of for fraud or non-fraud reason codes because we have the Dispute Product. We also know if you've requested a refund or return because of the Policy Product. That very rich level of data comes to us, and we checked, and it was actually three times what a payment processor might get. That all sits in a very unique modeling environment with supervised machine learning that's really been tagged and trained by us in our internal and kind of proprietary data. We focus on enterprises. We work with over 50 publicly traded companies. We're kind of best-of-breed solution in our space, they test us and continue to choose us. We've also created what we call an autonomous training environment.
If you're a large merchant working with Riskified, the model that we deploy for you is going to be trained based on your specific data points and tailored to you, which really just allows us to optimize performance to a very high degree. We don't need to have a very generalized model that works on all electronics merchants, right? We can really work with a single merchant that, again, we work with large complex enterprises, so we might have a different model on their gift cards. We might have a different model on their international travel business. We might have a different model on their domestic accommodation. We do all those things to really, again, optimize performance. That's the technology, really very, very broadly.
If we think about how the pricing structure works, we talked about the various components that make up the merchant's P&L, right? We do risk-adjusted pricing, and that's how we can work with very safe merchants in categories like groceries and food delivery and very high-risk merchants in categories like remittance and digital gift cards. If you're in a safer industry, we would provide you a lower fee because risk-adjusted, you have lower fraud rates. At the same time, we would also provide you a higher approval rate guarantee, right? Maybe a 10 basis point fee with a 98% approval rate. If you're in a higher-risk industry, luxury, fashion, globally, whatever it is, then you would anticipate a lower approval rate, 93%, and a higher fee, let's say 35 basis points.
In that sense, the risk-adjusted fee means that we're applicable to various kind of categories and industries, which we think is very important. The common thread and theme is that for 100% of our clients, we're able to outperform their internal systems and internal teams and drive value. Around the time of the IPO, we actually did an analysis, and for the top 10 clients, we were able to reduce cost by 30%. The fee that they paid Riskified was 30% lower than their cost of managing fraud internally. At the same time, we were also driving, I think, 7% incrementally higher approval rates.
That's helpful. Appreciate that. I guess given you guys have been around for over 10 years, how has fraud changed during that timeframe?
The great thing for us about fraud is that it's a game of cat and mouse. It's not something that you solve once and it goes away. In that sense, in order to go after it, merchants need to continuously invest resources in order to catch the newer fraud trends. Because fraud is becoming more sophisticated, it remains very challenging then to do that. Those challenges are better suited by a specialized company like ours to solve it. Why is fraud increasing? Just potentially the fact that there's so much more e-commerce and omnichannel, and you can get very high-value goods. The fact that there's no regulated market and law enforcement in these industries. Like if you were to go and call someone and say, "Hey, someone defrauded me in an online retailer for a few thousand dollars," they would say, "Well, I don't know.
Is that Korea or Russia? What exactly do you expect me to do there? The rise of AI tools and capabilities have also helped fraudsters perform more sophisticated fraud with a lower barrier of entry.
Awesome. You have talked about previously on earnings calls and at Ascent last year about the Riskified flywheel. Can you talk more about what that means and how it differentiates you from competitors in this space?
Yeah. As time goes by, we see that we get more accurate. We get more accurate because more merchants join the network and we have a better data advantage. We get more accurate because we're a bigger company and we're able to develop more machine learning-based capabilities that improve our engine. That just means that the value that merchants are getting from us is higher than it was when they started, right? They started at a 96% approval rate. Now they're at 99%. It's not just the approval rate. By the way, that delta in approval rate means that any competitor, it's much more difficult for them to reach our level and to provide value, which is why we have high retention rates. It's not just the approval rates and the core product. We've also added and started attaching these auxiliary products.
It used to be that Riskified only helped you with fraudulent chargebacks, but now we help you with your policy abuse issues, and that's significant, right? Most merchants that work with us on policy, we're able to block upwards of 10% of their refund and return requests and tell them, "That's actually, that's not a real customer. There's no one in Clark that never received a package. This is someone who created a fake account. We see them linked to others in our network, and you don't even need to honor this return request or refund request. You can just keep the money." That's massive ROI and savings for them right there.
Together with the dispute management product, the Account Secure, so we're providing more value to our clients as time goes by, helping us increase our kind of merchant size, our revenue base, helping us develop more products for them. I think we'll continue to see that type of product innovation and value creation for our clients.
Appreciate that. I've heard you recently mention your ability to resolve the true identity behind each online interaction with your Identity Engine. Can you walk through why this is specifically important?
Yeah. So when you think about policy as an example, what a retailer sees is, let's use the same example, right? Clark is a first-time new customer, and he's initiating, and he's saying that he never received his package. What do I do in this scenario? Probably 90% of merchants would say, "Reasonable. I want to give Clark an amazing experience. Let's honor that request." Because we're able to understand across our networks using various device fingerprinting and other links and kind of some more sophisticated cyber analysis, we're able to say, "Sure, this is Clark, but actually it's also Johnny from that store over here and Chett from that store over there." By understanding the identity, you can make a smarter decision. That's the same thing for the fraud decisioning as it is for the policy decisioning. For example, with our Dispute Resolve product, right?
A lot of the times the nuance is how to win what's called a friendly fraud, like chargeback, which is basically an underage kid taking their parent's card, buying something. The parent sees it, says, "I didn't order this," and then creating a chargeback. By being able to link the device or the identities in a household, we're able to submit that as file to the banks and have high win rates. Really understanding identities is the core of all the product platform that we have.
Interesting. As we think about AI commerce becoming the norm and Riskified's leadership in the traditional e-commerce space, what changes and how are you positioned to help customers navigate these changes?
I think it'll be interesting to see if kind of agentic commerce, where it goes and how it leads. It is a really fascinating use case. Will it be leveraged by the current existing brands, or is there going to be some kind of disintermediation by a centralized player where that's where you perform your shopping? Those are all interesting perspectives for us. The question is, how do we identify the person creating the purchase? It's usually related to bot activity, which we're good at identifying or not. Now that you've identified that it's a bot, it can actually be positive or negative, right? It could be a positive bot, and you want to enable that as a merchant, or it can be a nefarious, fraudulent, or even if it's not nefarious, maybe you just don't want that type of business on your store.
I think for us, it starts with being able to identify it. Once we're able to identify it, understand is the intent positive or negative? Will this result in a chargeback? From a go-to-market perspective, is this going to work through our existing channels, i.e., the merchants that we already work with, or is this one of the newer kind of AI platforms where the purchasing would happen through them? It is more of an integration and process with them.
Interesting. And then kind of pivoting here to the kind of customer base regarding the vertical mix and distribution of the business. Tickets and travel now accounts for roughly 1/3 of billings, which is a step up from where we saw 2020, 2021. How does the value proposition resonate with this specific vertical?
Yeah. So that's right. It used to be that about 45% of the business was fashion, predominantly luxury, and that's gone down as we've tried to diversify the base. Now we have 1/3 in fashion, 1/3 in tickets and travel, and the remaining 1/3 in other categories like payments and remittance and electronics and groceries and food delivery. I think what we've seen is that as we've been able to become more successful in a specific industry like tickets or travel, we're able to understand the nuances from a fraud and value proposition perspective, right? That helps us drive both kind of the incremental performance that they need. A, we obviously understand the fraud trends in the ticketing space related to brokers or fan-to-fan purchases or primary tickets or secondary broker sales.
Those are all nuances that we know how to take into account that help us drive incremental approval rates. At the same time, we also have the best network effect within these categories that helps us drive superior performance. We are also able to come to every secondary or primary ticket and say, "Hey, look at this list. This is everyone who works with us. Don't you kind of want to join as well?" Which are all helpful, right? We are making a conscious effort to expand into newer categories for us. I think we highlighted payments and remittance as one that we feel has a lot of promise. We think we will continue to see that trend of being able to build a network effect within some of these categories.
You also call that international growth being strong in recent quarters. Can you talk about the company's go-to-market motion outside the United States?
Sure. I think we shared that eight out of the 10 top clients in Q1 were outside the U.S. Very, very strong international growth. We have global direct sales in a lot of regions, and our sales team is by geography and merchant tier. What we see is that once we are able to build a presence and gain a foothold with one to two marquee accounts, that helps us penetrate and build a wider pipeline and kind of take, like we do from an industry perspective, also gain more market share in a geography.
Awesome. One of the most more interesting components, I believe in my view, is that the non-Chargeback Guarantee products and kind of the opportunity set that you have there. Can you talk more about these offerings, the problems that these offerings solve, and what really these products kind of bring that's incremental to the value proposition?
Yeah. You're right. The newer products grew 190%, and they generated about, year over year, Q1, those are the results we shared. They were about $4 million in 2024, and we anticipate them to be low double-digit millions in 2025. The biggest revenue component there is from our Policy Product. Our Policy Product, like we discussed, helps merchants analyze returns and refunds and tells them, "Stop. This is abuse. You do not need to honor that." More than that, it allows them to build a different experience for different customers. Some merchants are leveraging this product to say, not just, "Hey, Clark initiated a return. Do I provide him a refund or not?" It actually goes a step further and says, "You know what?
If Clark is one of my best customers, and they can define best customers based on LTV, propensity to buy, there is a lot of functionality there. I want to offer Clark, once he hits that refund button, give him the cash immediately or give him the opportunity to select a new size and color and ship it to him directly. If not, maybe I will wait until you drop the box off at UPS or UPS collects it and you scan it in, and then we know that reduces their likelihood of fraud. If I do not know Clark at all or I think he is not one of my best customers, I am going to wait until the package reaches my warehouse fulfillment system.
I've inspected the goods that you're returning, and only after I've inspected them, I issue you your refund." Basically, the way to think about it is if today a lot of merchants are doing something very generic and unsophisticated in the sense that every customer receives the same experience, merchants are using our Policy Product to create different flows and experiences based on the type of customer. I think that's really exciting.
With the new product cycle, what are your expectations between growth and the formula going forward between new logos and expansion?
Yeah. I think we've always seen kind of a healthy mix of new logo growth versus upsell and cross-sell to the existing base. It's ranged from 50/50 - 60/40 in both directions depending on the year. I would anticipate that to kind of continue from what I see.
Got it. From an expense perspective, discipline has been a focal point in recent quarters. How do you think about balancing investments in the future to continue to drive growth in the top line?
You're right. Over the past three years, we've really focused on not just flattening our expense base, also lowering it, and a lot of great accomplishments there. Basically, every gross profit dollar has flowed through to the bottom line. I think we shared some kind of early indications about maintaining control for 2026. Beyond that, at some point, I do think it would be incrementally beneficial to continue to invest at some level.
Got it. You also planned, I think it was in Q4, that you mentioned expanding the R and D capacity. What areas are you most excited about from that standpoint?
We continue to see just a lot of traction in the newer products. And it's not just the revenue opportunity themselves, because when you think about the growth in the pipeline and the overall competitive win rates, which have been at or above 70% past few quarters and have improved, a lot of that is because of that platform sale, right? So we're basically providing guaranteed ROI. So why aren't more merchants joining faster?
Some of the pushback has always been, "This sounds interesting, but I have capacity to integrate four new projects in a year, and I'm changing my finance system, and I have a new marketing stack, and chargebacks isn't a burning priority for the business right now." Now that our value proposition is wider and we're helping solve more problems, we're able to both generate more ad bats and get more resources from the clients to integrate the solution, which is just a longer way to say we believe that by continuing to invest in these R and D capabilities, it would help drive more growth as well.
Got it. Is there anything on the product roadmap that really excites you that you could potentially elaborate on?
Outside of the products that we've discussed, there's nothing that we've announced publicly and just from a competitive situation perspective. I think that's important. We do see ourselves continuing to innovate. I think that the runway in the products that we have announced and also in the core Chargeback Guarantee is still significant, right? If you think about e-commerce being the trillion-plus opportunity that it is, we're proud of the $140 billion in GMV, but there's still obviously a lot of opportunity to grow. As we think about creating more value through both the chargeback product and the other products, we think it'll be easier to add more GMV to our system.
Appreciate that. In terms of the macro, how is Riskified positioned to weather a more volatile environment? What impact, if any, is there from the implementation of tariffs?
Look, historically, macro has impacted us. Whether that has been as we have come out of COVID, stay-at-home beneficiaries like Peloton and others have had meaningful volume drops. We share that originally 45% of the business was tied to fashion, some of that luxury fashion. That has been a headwind and is still showing softness. Because of that, we have really tried to diversify the client base and really focused on creating something that more resembles the wider e-commerce environment. Some of the recent adds into the non-discretionary categories are food delivery and groceries. Food delivery, obviously, is that discretionary or not? That $24 burrito post all the different fees, that is another question. Remittance and payments as well. I think we are closer to approximating the wider e-commerce environment, which we would call as high single-digit growth.
At the same time, our macro same-store sales have been flat to slightly negative over the past few years.
Thank you for that.
Sorry, for tariffs. I forgot the last part. We sized the direct impact of China to U.S. as less than 1%. We think we're not very exposed there.
Got it. Got it. I guess also year to date, you've talked about the fact that the pipeline has been very strong. Do you see more growth coming from, and this is, I guess, alludes to kind of the growth mix, is that coming from new logos? Or how is growth being driven by expansion of sales to existing clients? To your point about same-store sales.
No, the pipeline is either coming from, obviously, from, so same-store sales increase would not be in that category, right? We would look at it separately. We would say the pipeline is being driven primarily by new logos and upsells, meaning capturing more volume from our existing clients and to a smaller degree by cross-selling our newer products to the existing base.
Appreciate that. Moving more to the technical side and the perspective that you provided maybe initially here, how has the platform changed? Can you talk about the data advantage that Riskified has over maybe said peers and competitors in the space?
Yeah. I think the depth of the data that we shared, like the amount of data we have per transaction, not to repeat myself, but that's probably a key component. I will say that since day one, we've always done the Chargeback Guarantee. What's unique about it is that in order to get reimbursed, merchants provide their chargebacks to us, meaning that we continuously have source of truth tagging and data to train our machine learning models. The first two years of Riskified's existence, we were basically getting high-risk transactions that other systems were declining. We were receiving those transactions in order to make any money. We had to approve that. Obviously, we need to be very certain because we need to pay back if we make a mistake. We were somewhat manual at that point, right?
We would do a lot of manual checks and reviews and try to understand how to approve that transaction. Through that very slow and meticulous process where we own that data and training set, that's how we build our automation and machine learning systems on top of that. I think that's just another one of the reasons why the scaled automated engine that we have right now started from very beginnings that are difficult to replicate at scale with a high barrier of entry.
Appreciate that. With that advantage, you also kind of flipping back to kind of the financial framework here, it is able to scale effectively, and there is a lot of operating leverage in that model. The target that you outlined was 15%-20% EBITDA margins in 2026, which implies that we should see a continued meaningful expansion from current levels. What are the main drivers to get there, and does that include top-line reacceleration?
Yeah. We kind of talked about that 15% kind of north star. We want to make sure that we manage expenses until we reach that trajectory. We definitely anticipate an acceleration from current levels exiting the year to kind of double-digit growth rates, which we think would be a meaningful driver there. Some of that is more mechanical. We disclosed kind of a large client churn the prior Q4 2024 that we sized in the range of $20 million for 2025. Just by virtue of flapping that, that's kind of incrementally accretive there. Being able to maintain the leverage that we've shown over the past few years should be kind of good drivers.
Appreciate that. Also in terms of the financials, you already have a stockpile of cash, a pristine balance sheet with no debt, and you've already been actively repurchasing shares. Do you see this as the best use of capital at this point in time?
We do. We look. We try to find the right M&A opportunity, and that for us would be an interesting technology that we can cross-sell to our existing kind of blue chip base that looks at us as kind of great partners. Unfortunately, I think the bid-ask spread between public companies, private companies, or current valuation is still somewhat large. Within that analysis, we have historically always defaulted to let's just increase and authorize more buybacks.
The philosophy around M&A that you're using, you haven't necessarily been inquisitive, but could you walk through how you view shareholder value and what would excite you from an M&A perspective?
I mean, math. If you can buy it for $10 million and you think that you can cross-sell and generate $30 million in a short period of time, that would excite me. At the same time, if we believe that we can take out shares today and that the value based on kind of future cash flows that we anticipate over the next few years or quarters would be so, so that excites us as well. I think it's very kind of based in financial analysis.
Appreciate that. As of December 2024, Riskified, you reviewed more than $140 billion in e-commerce and GMV globally, but the total e-commerce GMV is over $6 trillion. How do you go from where you are today to becoming a larger market shareholder?
Yeah. Look, I mean, obviously, we do not think that the $6 trillion is what we consider our SAM today. Even if you back out some of the things like China in and some Amazon volume and some PSD2 regulation, you are still left with a multi-trillion dollar opportunity, which is incredibly exciting. For us, it is how do we continue to create more value to clients? Remember, a few years ago, a fraud manager was using us to save on chargebacks and approval rates. Now, by the way, more people within the organization, the fraud manager, the support team, sometimes the security team, they are using multiple products and getting so much more value. I think that allows us to capture more GMV, and that in turn helps us become more accurate and provide more value, which should in turn help us accelerate.
I think it's that type of kind of product-led value creation together with good go-to-market distribution strategy, which would help us capture more.
Got it. You mentioned that enterprises try to do this in-house. Why do enterprise merchants need Riskified?
It would improve their P&L. They would be able to pay less and get more.
Appreciate that. Simple and concise. Maybe lastly here, in terms of the long-term time horizon, where do you see Riskified down the road and where do you want to be in the next five years?
Look, I think we would have succeeded, and again, not getting into guidance and numbers, but if we continue to execute well on our plan of expanding the product portfolio and generating more value to our clients and leveraging that to gain more capabilities and more geographies and more industries and creating those network effects, I think that would be success.
Awesome. I appreciate it. Eido, does anyone have any questions?
Yeah. So who would you consider to be your closest competitor? How is your product different from others?
We tend to define our closest competitors as in-house teams that leverage numerous solutions, right? Those solutions could be kind of the legacy solutions that some of the card networks have to a handful of some more newer and modern. Really, the main decision point is, do I want to continue to invest and build this internally with cost involved? Different people in the organization have different priorities there. If you ask the CFO, he's going to be like, "Well, no, why?" If you ask maybe the person managing that team, he would have a different perspective. That is really the main decision point that results, "Okay, I'm going to go with Riskified or not.
Can I ask a quick question? I don't know if I've laid this story way, but you talk about the risk-adjusted charge to the enterprise to your client. So if you're efficient in that, is the margin on each of those clients similar?
Yeah. Overall.
Would you rather have a client that you get, right, that you're charging more, but there's a reason you're charging more? Is that a better term?
Yeah. So we have what we call the CTB ratios, right? The chargeback to billings ratio per client. On an aggregate basis, obviously, everyone is profitable. The difference are for very high, for a client where you take a high fee, let's say 100 basis points, you should be willing to operate at a higher chargeback rate and only have 20 and have an 80% chargeback to billing ratio because that 20%, that's a lot on the net spread, right? For someone who you take a 10 basis point fee, then potentially you would want to make sure that your margins are higher there because the spread is much lower. That's how we think about it. By the way, that's why we say take rate is an output of the business and not really like a model going in. It does not tell you a lot.
Like take rate went down, it could be great because from a net perspective, we just got a massive client that's on like a shop protection deal where we get everything. That could actually be better than having the same client give us 10% of their risky volume for a much higher risk-adjusted fee. Take rate would be higher in that scenario, but net revenue to us would actually be lower.
Awesome. Thank you, Eido. Appreciate the time.
Appreciate it. Thanks.