All right, fantastic. Welcome back, everyone. We are here with Eido Gal, Chairman and CEO of Riskified. Thank you so much for being here. I really appreciate it.
Thanks for inviting me.
I think likely most people in the room are familiar with Riskified, but maybe you can give us a brief overview of the company.
We started as an AI solution to help e-commerce merchants manage online payments fraud. We have expanded over the past three years to help them with their various policy abuse problems, chargeback management, and account security.
OK. Maybe kind of following that thread a little bit, describe the sort of merchant challenges that Riskified solves. Why do merchants need your solutions? Maybe after that, I want to follow on with the dimension of the problem that they're trying to solve. What is the core value proposition?
Sure. When merchants accept online credit card transactions, they're liable in the event that they've accepted a transaction from a stolen Credit Card. They need to pay back the customer. That's a really challenging problem that involves some elements of CyberSecurity, big data, machine learning. They have large internal processes and teams that try to identify good transactions from bad transactions. That's on the payment fraud side. On the policy side, they also get a lot of requests around, hey, I never received my package, or I ordered a red pair of shoes, and you sent me a black pair, refund and return request. A lot of those are actually abusive. They have all these different areas and vectors where they can receive loss. For example, on the chargeback side, there's a lot of what in the industry is called liar buyer or friendly fraud.
I actually order something, and then maybe it's a bit too expensive, or I didn't like it, so I just say, file a claim, a chargeback claim. All our products really help within that ecosystem, from identifying good transactions and accepting them towards kind of helping them reject fraudulent policy requests, towards managing the dispute management process with the banks.
This might be another way to ask the same question, but I want to make it a little more palpable for our audience. How do your customers, your successful customers, talk about the impact your products have had on their business, especially in ways that maybe go beyond the usual metrics like chargebacks and approval rates?
Before we go beyond those, let's start with those because it is an important aspect. Around the time of our IPO a few years ago, we did a large analysis. From our top 10 customers, the reduction in cost that stemmed from using Riskified was upwards of 30%. The average incremental approval rate was, I think, an increase of about 8% or so. Very meaningful ROI there. What we see is that we also help the customer experience. On the recent earnings, we share that we've seen a lot of growth in the payments and remittance industry and category. There we're not just helping them kind of approve more transactions, reduce the overall cost, but overall improve the customer experience. You can send out more funds in a shorter time window than they might have enabled with a less sophisticated fraud vendor.
When you think about kind of instant ACH versus waiting for a four-day rolling withdrawal window, we help improve that as well. It is also helping the customer experience in a lot of these verticals, while at the same time improving acceptance rates and reducing the overall cost of ownership.
Talk about your kind of go-to-market motion. How do you go about attacking this sort of large problem that exists out there?
We really focus so far on the enterprise space. We believe that enterprises, they tend to choose deep best-of-breed solutions, and they have the teams and the capacity to test and kind of integrate those. We believe that's what we are, the best-of-breed solution to looking at e-commerce transactions, understanding if they're fraudulent or not, or helping you manage your various policies. We have an enterprise sales team that targets accounts from anywhere from $50 million to $1 billion. We also serve, but really the sweet spot is $1 billion plus, work with over 50 publicly traded companies. That team goes after them directly. We've kind of had a lot of growth and success in some new regions for us in Latam and in APAC, have regional offices there in China, Australia, Brazil, Mexico.
Probably mid-market in SMB distribution is the next frontier for us that I anticipate we'll attack over the upcoming quarters.
One of the things I think that's interesting about the seat that you sit in is that you've had sort of a bird's eye view to the evolution of fraud and so over the years. Maybe taking a step back, just talk about how the types of fraud that you've seen have evolved over that time frame.
Fraud is interesting, by the way. It changes quickly and dynamically. It is this constant race to stay one step ahead of the fraudsters. I would say that we have seen, especially in the past kind of one to two quarters, an increase in sophisticated fraud. That could be anything from fraudsters leveraging kind of new AI capabilities to talk to different chat agents and bots on the merchant's website to gain access to accounts and leverage that to perform fraud. We have seen an increase in remote desktop. When people take over a good device, they send you a link, and you click on it, and now they have access to your computer. They leverage that to perform transactions. There has definitely been a marked increase there. Overall, we think that is a positive for us. That is a positive for a few reasons.
Number one, it means that merchants on a standalone basis have a hard time understanding and combating this type of sophisticated fraud. We obviously have a much more sophisticated R&D team to identify and solve those problems. Once we see it on an individual merchant, we can also plug that problem for other merchants on our network. We think overall that's strategically important. We've also mentioned that our pipeline has been growing and is really at or close to record levels. Part of that has been because of the sophisticated fraud, there's been an increase in losses for merchants driving them towards our solution.
There is sort of a built-in cyclical driver, not cyclical, like a fraud cycle almost that is a driver for your business. The fraud changes, the merchants need the latest and greatest to combat it, and you are there to provide it, basically. Is that a?
I think that's a fair assessment, yeah. I mean, I think over the time that we've seen, fraud has increased in attempts and sophistication versus seen a decrease. There can be various reasons. If you think about e-commerce several years ago, it was not tiny, but it was small relative to the $6.5 trillion today. It's a very large opportunity for fraudsters. More and more industries are coming online. Merchants are trying to create very simple omnichannel frictionless flows. That's also very challenging from a risk perspective, inviting fraud. Obviously, there's little to no law enforcement in these areas because of jurisdiction in different geographies and just even understanding who's performing the fraud and where they sit. It's really easy to buy sophisticated tools that will enable you to perform fraud on the dark web. I don't need to be a sophisticated attacker or hacker.
I can just buy a service for $30 on the dark web and use that to perform fraud. I think there's a lot of tailwinds for the fraud industry, if you will. And yeah, we're seeing that.
I wanted to just revisit something that you mentioned, too. That is the use of AI by the Fraudsters. I want to get into in a second how you guys are using AI in your own business, maybe to combat the fraudsters and for your own operations. From the fraudsters' perspective, does AI represent a new killer app? Are you seeing more? Do you think the emergence of that should raise alarm bells with folks?
I mean, I believe that part of the increase in the sophisticated fraud that we're seeing is also driven by adoption of the more sophisticated AI tools and capabilities on the fraudster side, 100%.
I see. Maybe then flipping that the other way around, you guys were a little bit ahead of this Hype Cycle with AI, and that was always a part of your business. Maybe you can talk about how you use AI and machine learning. That might be on the operational side to improve the efficiency of your company. It might be methodologically to combat fraud. How do you use AI to combat the fraudsters using AI?
That's a great point in question. Thank you for bringing it up. I think that in early 2013, when we started the company, unfortunately, we called it machine learning and not AI back then, but basically the same thing.
Used to be a Payments Analyst. Now I'm a Fintech A nalyst.
Fintech and evolution, traditional evolution. We definitely started there. We think we have a pretty meaningful head start in developing our models and having a very proprietary data set of good transactions and bad transactions to train our supervised machine learning on. I think one of the more important things that are happening now because of this cycle and the attention and the focus is that we have the ability to look at everyone and say, look, this is not open research AI. This is something that is very targeted to solving a meaningful business problem. That is exactly what you are looking for right now with modern AI solutions. It has been purpose-built for that. It has its own proprietary data and tagging and modeling techniques and capabilities.
The nice thing is that even as you think about whether it's macro challenges and merchants trying to focus on reducing their expense base, if there's any area that's growing, it's AI tools. Hopefully, we can also be a beneficiary of that.
I wanted to ask you about the macro environment. I think you mentioned on the earnings call, for example, with tariffs, it was relatively minimal impact. That has kind of been a theme at the conference here, just checking with various folks on what you're seeing. Maybe from a macro perspective, a Tariffs-related perspective, what are you seeing out there in terms of your flows?
We did not see, and we saw a resilient consumer year to date at the time of kind of the quarterly earnings, the things that we had anticipated we would see, which are slightly lower luxury fashion trends, but still improving. We saw robust travel activity, more intra-Europe and other verticals than kind of international U.S. travel. Overall, encouraged by what we saw. Obviously, still high degree of uncertainty in the back half of the year. Internally, I think we sized the potential exposure to China-U.S. as less than 1% of volume.
Your view, I presume, is more of kind of a non-discretionary read on non-discretionary spend. Is that fair? Or do you also handle discretionary spend in your flows?
No, I think we have a lot of discretion. It's true that.
Reverse that. I meant to say discretionary versus non-discretionary.
Historically, we've had a lot of discretionary. The vast majority was discretionary. We've actually and relatively purposefully have been targeting non-discretionary categories. A lot of the growth that we've experienced in verticals like food delivery, groceries, payments and remittance has really created a more diversified base that also has a more significant mix of non-discretionary relative to prior years.
OK. I want to ask about your net dollar retention, which I think there's a macro overlay here. I think you've seen some declines in net dollar retention rate. I think it was about 96% in 2024, about 110% a couple of years before when the environment was different. What do you think is the driver there? Is that a macro? Is that a function of the Macro Environment?
It's also a function of the macro, but it's not just. I think there are two big reasons. Number one is, unfortunately, we disclosed a large merchant churn a few quarters ago. That's kind of a Q4 to Q4 impact. I think we sized it at anywhere from $18 million-$20 million for calendar 2025. That's also impacted in 2024 and is impacting the numbers there. On a positive note, we've kind of shared that I've had success at renewing 100% of top renewals over the past two quarters and are projecting continued success in that area. On the discounting side, we've been able to manage to less than 10% of renewals to receive a discount. We've been really leveraging the product platform to achieve that. I think those are areas that have impacted us and will show improvement just based on the team's execution.
The second area is the macro, where some of the areas, especially the more discretionary categories, have been challenged over the past few quarters relative to what we've saw historically. There are, I think, two things. One is that we absolutely believe that that will revert to the norm at some point over the upcoming few quarters. The second thing is, I think that our diversification into non-discretionary categories will overall help us mimic just the overall e-commerce growth rate a bit better.
I see. Some macro, but some idiosyncratic factors as well. This is a kind of a question about the competitive environment. Who are you competing with? Are there fintechs who are doing something similar to you? Is it in-house solutions? What does it mean exactly? How would you define a competitive kind of set that you go up against?
Sure. I think when people think about competition, they can think about some of the more traditional legacy players in the space that are dedicated risk and fraud solutions. They have the majority of the market. They're probably not gaining a lot of new clients, maybe retaining some of them over some renewal cycles. Overall, they're probably losing more market share. These are technologies from kind of 20-plus years ago and really not AI-based and challenging for them to move in that direction. You have the more traditional kind of gateway risk solutions, whether it's from the newer modern payment stacks or even some of the more legacy stacks. I think there they have a few inherent challenges, especially as we focus on the enterprise side. One of the first is they have a limited amount of data capture relative to a solution like ours.
That limited data capture results in suboptimal modeling performance. When you think about Riskified, we actually see a transaction from the time a customer is browsing on the merchant's website, creating an account, putting something into their cart, the checkout flow. We also get information if you contact the customer support team because you want to change delivery address. We also get information around your refund and return request. We get information on any type of chargebacks that you filed. We have that data across our network. Sorry, I should also add, we also have merchant-level specific information. For example, if we would work with an airline or an OTA, we would have info about your passport, the cabin affair that you're doing, and the flight route because it has some risk information.
When you compare that level of data, how we can create features with it to the data that a kind of payment gateway captures, it is really in parallel. I think there is an inherent advantage to a dedicated solution with the data capture that we have. The second thing is that our clients tend to be sophisticated, have payments teams, and they run a multi-acquirer setup. They do not like to have tie-in to a single vendor, especially on the payment side. They like to do their own vault and orchestration. I have seen very limited uptake for these types of solutions in that space. That is why we think we are better protected in that area. Probably the last category is modern dedicated risk solutions, which I think there are a handful of smaller ones in the market today, not just us.
I will say that I think our competitive win rates have been increasing pretty nicely over the past few quarters to over 70%. For us, the biggest focus is what do we do to increase the pipeline? What's been working well for us in that area has been the expanded product platform.
Maybe that's a good segue to talking about the pipeline as you see it now and as it sort of will feed into the growth remainder of the year and maybe into next year. How's the pipeline looking, relatively speaking?
At or close to all-time high right now. I think three reasons driving that. I think the first and most important reason is the product platform. It just means then instead of saying, hey, Ramsey, I can help you with your chargeback problem. You have 30 basis points. I can save you, and you'll have 25 basis points. You might say, the integration is complex. You do need a lot of data. I have four other priorities. Let's talk again in a year. Maybe I'll have integration resource. I can now tell you, I can also save you on the policy abuse issue that you're solving. That's top of mind for a lot of merchants. I can help you optimize returns and refunds. I can help you manage the entire dispute management flow with kind of higher win rates. I can help you secure your accounts.
Now you're saying, oh, that's actually interesting. That ties into a lot of different strategic initiatives that we have. Even if I'm not going to start with the entire platform, I can actually integrate you now to solve that use case. I think that's really been helping in generating more at-bats. The other two things that I would mention are, number two, the increased fraud pressure that merchants in the environment have been seeing that we highlighted earlier. Number three is just some of the go-to-market geographical diversification that we've made and have kind of robust teams in Brazil, in Australia, and China. That's been helping generate leads and deals in those regions.
It sounds like, as you mentioned, the new products are helping with signing your logos. Talk about the customer journey. Once you sign the customer, do they sign up initially for a lot of different products? Is there a longer-term upsell, cross-sell type of an opportunity, gaining share of wallet over time? How has that looked over the past few years?
I'll differentiate between our existing clients where it's much more of a cross-sell opportunity because historically, we were a single product company. There just wasn't an opportunity. Now it's much more cross-sell towards the newer clients, which we've had some great examples of some of. I think we had one recently that went live with the entire platform day one. That's something unique. We're starting to see more of that. Probably the norm is still to start with one or two and expand over time.
Having said that, over time, that share of wallet is something that you are seeing an expansion of?
For sure. When you think about, I mean, the growth in new products has been significant. They have gone from kind of mid to low single-digit millions to projected low double-digit millions. Q1 revenue growth from them has increased 190%. We do feel that they are gaining meaningful traction. The value that they are providing to merchants, which is probably kind of the best indicator that I can see, is significant. At the same time, we are also expanding Chargeback Guarantee volume with our merchants. On the one hand, we are proud of the $140 billion that we have guaranteed in 2024. That is still kind of relatively small compared to the size of the overall opportunity. I think the platform will also help us just kind of gain more chargeback volume as well.
On the guarantee product or offering, who is that tailored for? Who is interested in that? Maybe overlay that product with your client base. Who is that product for versus maybe folks who opt to use you but maybe do not take that solution?
Yeah. Just to level set with everyone listening, the guarantee basically means that we tell merchants, listen, your cost basis is 30 basis points and your approval rate is 90%. We'll guarantee you a 92% approval rate and a 24 basis point cost of fraud. We'll approve those transactions. If we make a mistake and you receive a chargeback, we would pay you back. That is different than what can sometimes be called a scoring tool or an in-house model where they need to make the decision, but they're also ultimately liable. Or they can trust someone else's decision. The reason we do that model is very simply, we believe that it provides more value to the merchant. We've had instances where merchants have told us, listen, I came into the office.
Over the weekend, there was a model drift because of a bad score input from one of my vendors. I just lost $5 million. I do not want that. When people do not have, when the fraud vendor does not have accountability over their performance and decisions, the merchant needs to continuously manage that and have a team and have different kind of vectors to look at. For us, it is really a way to leverage the sophisticated technology that we have created in order to create a better experience for the merchants and more value. It does not matter what industry they are in. It could be like a safe industry, a high-risk industry. Providing guaranteed performance would always provide more ROI. Really, the only difference is pricing. When you think about traditionally safer industries like domestic U.S., kind of groceries, probably lower risk than Global Luxury Fashion.
We have Risk-adjusted or Risk-based pricing, but the value is there for both of them.
Talk about an aspect of that business is basically the calculus you have to do on the back end to ensure that you're structuring the deal correctly, that you don't get caught in a situation where you're absorbing chargebacks more than you anticipated. Looking back at the I've covered you for a few years now. That doesn't seem to rear up as a problem. Just curious, talk about the sort of the risk control and underwriting kind of process of how you frame that up.
For an industry or a geography where we have experience in, we're probably going to do a great job right off the bat. Both reduce the merchant's overall cost basis and have very healthy margins or what we call our CTB chargeback to billings ratio. In newer regions or newer categories, we might start with a higher chargeback ratio and improve that over time. What we've seen consistently over the history of Riskified and the state today, in virtually 100% of the cases, we're able to outperform a merchant's existing stack team risk solution, not just outperform that by a bit, outperform by such a significant margin that we also have kind of a healthy profitability in that area, which I think just speaks to the quality of the technology and the accuracy of the decision.
Earlier in the year, you announced Adaptive Checkout, which is, I think, a configuration of Chargeback Guarantee or at least a fraud and risk engine. Give us an update on how, maybe tell us what that product is and also give us an update on how the rollout's going.
When we were talking to merchants, we were just talking about their merchant acceptance rate, a transaction that came through kind of the payment funnel, the gateway to the issuer, to the acquirer, to the card network. The merchant has the ability to kind of often capture in technical terms. We saw that risk is so embedded into various decisions in that overall approval matrix that we can create a higher overall conversion experience for our merchants. This is in consultation talking to them. That thinking led us to develop Adaptive Checkout. What Adaptive Checkout actually does is pre-auth. It sends enriched information to participating issuing banks like Capital One, like Bank of America, like Discover Card. We are adding more and more. By having this enriched information, the issuing bank can have higher approval rates for our e-Commerce merchants.
That rich data helps them make smarter decisions. We also screen out bad transactions from even entering kind of the payment flow. Our merchants experience a better kind of trust score with the issuing banks and everyone. It's not like, oh my god, they're sending so much fraud. This is a terror. It's not someone actually makes that decision. There are models on the card network and the issuing bank side that say, well, if there's a ton of fraud coming in, there's an issue here. We might want to reduce approval rates. We do that as well. We also have the ability to trigger various very targeted flows like request SMS confirmation, do a 3D Secure check, do a one-time password authentication, request CVV if we're kind of afraid that there might be an ATO in this scenario.
It enables us to create very targeted and customized friction when the overall goal is improving conversion rates end to end and not just on the merchant acceptance side. It's been gaining great traction because it's bundled with our Chargeback Guarantee product. It's the ideal integration that we recommend. It's been generating in the area of 100-250 basis points improvement in the end-to-end conversion rates.
You talked earlier too about some other newer products. I'll say newer and not new because they've been around a little bit. Maybe you could also kind of give us a brief description and an update on Policy Protect and Dispute Resolve.
Yeah. Policy has a handful of different use cases. Merchants are using it to manage their business logic. They go into day-to-day to kind of create different rules and flows. Basically, the main category is, number one, to prevent abusive refund and return requests. That means that I do not need to steal your Credit Card. I can just order something online, call the retailer, say I never received my package, give me a refund. Fake accounts, new identities, stuff that our fraud engine is very good at finding. We have been able to block upwards of 10% of refund and return requests with no incremental increase in false positives. Merchants love this, helping their profitability significantly. The second value is more around who are your best customers and how do you give them a better experience. We just talked about blocking the bad guys.
If you're a great customer, let's give you the refund instantaneously once you say, I want to return this. Some merchants would say, we want to wait until the package returns and we inspect it. For my best customers, I'm going to give them instantly. We have merchants that do this to provide different shipping options for various tiers of customers. Maybe my best international customers will receive free returns, and everyone else would have a restocking fee or something like that. Merchants are using it as they do kind of product launches and item limits. If we think about a sneaker launch, a limited inventory for ticketing or concerts, how do I ensure that real good customers receive that inventory? It's not a bunch of bots buying it and just making all the money.
We have some sad kids somewhere who were not able to participate in the launch that they waited all night for. The brand experience there is important. Those are just some of the use cases for the policy product. It has been gaining very meaningful traction. On the chargeback management side, it has been a great workflow tool for fraud payment support teams to manage the representment process. It can be configured to be done 100% automatically. It can be configured to be done semi-manually if you want to add specific data points for high-amount transactions. There is also an AI component here when we say, hey, we think we understand now how to optimize the dispute process versus specific issuers' card networks to optimize the win rates there. I think it is really important for us to become more of a workflow.
As a management team, we sometimes host clients and listen to them for their feedback on our product. One of them recently said, we asked them what product, and they're using the entire stack, provides the most ROI to the business. They said, it's Chargeback Guarantee. You've reduced our cost. You've increased our approval rates. That's like the business perspective. That's the highest ROI. When I think about my team, and this is a payment manager that has like 10 to 15 agents working for him, it's the dispute management tool because it's like our day-to-day tool that we use to manage this process. It's kind of miles ahead than anything else we had. We love that being more integrated into the workflow as well.
Fantastic. I think we're out of time. Great conversation as you go, Eido. Thanks so much for joining us.
Thanks. I really appreciate it.
Yeah. Thank you.