Okay, good morning. Good morning. Welcome again to day two of the DB Tech Conference. My name is Brian Keane. I cover the payments, processors, and IT services here at Deutsche Bank. We're really excited to have Ed McLaughlin, who's the President and Chief Technology Officer at Mastercard, here to do a fireside chat. Nothing going on in technology these days, so I'm sure there's something to talk about. With that, Ed, maybe we could just kick it off. As the Chief Technology Officer, can you tell us a little bit about your background?
Sure.
I know you've been at Mastercard for a while, and discuss your current role, what you're working on now?
All right. I do want to ask one thing for everyone in the room, though. For everyone who's not here, tell them they missed the best session. All right? Will you do that, will you do that for me? So, quickly, and let me just do two things, because you asked about the background. I do think it's interesting. You know, I have always worked at really the intersection of technology and data, transaction processing. The first product I had in the late 1980s was what we called an EDI translator, where we were linking together isolated computer systems to move data and the associated payments with that. And there's, like, 10X, 100X value you can generate by doing it, and you just learn that early on.
So having been able to watch what was futuristic become the commonplace, which I know you have all done as investors, has just been amazing. I was in packaged software. We had a company, Logic Works, that made data modeling and design tools for the first wave of data warehousing, relational database, separating data from applications to generate value from that. We IPO'd that one in 1996, sold it to Platinum Technology, now CA, now Broadcom. It's hard to keep track, I'm sure you have the same problem. And then I was CEO and cofounder of a first-wave internet payments company called Paytrust. It was online bill pay applications. Sold that to Metavante in 2002, and then ran the payments and strategy for the Metavante group before I came over to Mastercard in 2005.
I've been at Mastercard about 20 years, but prior to that, it was mostly what we didn't know to call fintech at the time. I think it was e-finance, but technology startup software. At Mastercard, I'll do this quickly. I really focused on how do we expand where we could use our applications, taking advantage of new flows and new segments that we could pursue. I had healthcare, bill pay, transit, which has been just a home run for us, and more and more ways we could apply the technologies we had. Ran the global franchise, including through 2008, kind of a crucible moment for a lot of us in the downturn, and did the digital and emerging. This was really when the shift to digital first to mobile was happening.
So we did things like MDES, which is our tokenization system, Send, which became Mastercard Move, moving, you know, linking accounts together faster. We worked with Apple Pay and Square and Uber and Netflix and that whole generation as we were able to transform, I think, in many ways, what payments meant. And twenty sixteen, went into the CIO role, twenty seventeen, CTO, President and CTO for Mastercard, running the tech. Really quickly there, you know, the operations is who we are and what we do every day, keeping our promises, what the brand means, and that trust we have in the brand, and then also all the technology to build what we become. So the investments in the modernization, the future of what Mastercard and hopefully the industry becomes. So it's been great.
Yeah, no, it's been a great ride, and it sounds, feels like there's a lot more to come here.
Oh, yeah.
So maybe you could-
Getting started.
Yeah. Maybe you could just talk about innovation at Mastercard and share some of the investments you guys are making in new tech to modernize the network?
Sure, and I'll start with the network, because we've made some amazing investments over the last couple of years. We even won our own CEO award for this one, of what we've done to enhance and upgrade the network backbone. So if you think of Mastercard, really what we're running is a global network fabric, incredibly high speed, incredibly resilient, with both our own data centers, public and or private and classic cloud, as well as working with the third-party clouds as peers within that network fabric. So it's giving us the ability to put compute, to put data, to put servicing anywhere we need to in the globe for optimization or localization or regulation. So we've really built a tremendous amount of flexibility into the network. Now, that helps us to improve resiliency, lower costs, speed everything up.
But what it also gives us is an incredible foundation to build on top of. So one of the biggest things in the last few years is what we call our Cloud Connector or our Cloud Edge. We want our customers to have a single front door for transactions, for services, for APIs, and we'll meet them wherever they are. So we've always had edge computing we could put right in our customer's data center for unbelievable reliability and security. We've now recreated that when they run in any of the major cloud providers and regions. So whether it's a traditional customer we've worked with for years that's shifting some of their important loads into a cloud environment or born in the cloud fintechs like, Nubank, which is a great Mastercard customer, we can connect to them directly that way.
Second big element of modernization is we've really deconstructed a lot of the services that used to just be part of our transaction flow, things like tokenization, things like the fraud scoring, loyalty services, installment payments, and provide direct access to that now. So you can either get it by running a transaction through our network or, for private label or other reasons, we have the ability to flexibly provide those services to transactions that we're not switching ourselves. That's been transformative, really transformative for us. The other one I'd highlight is just the virtualization of everything. Our ability to take what used to be relatively fixed or physical things and drive it on a truly virtualized basis allows us to work with things like tokenized deposits and accounts, to creating new, what we call contextual commerce...
new ways that you can access the network, new ways that consumers are going to want to interact. One example, just because I love it, is we've always talked about your car being a computer on wheels. So we're working with Mercedes, for example, where we can virtualize the connection to the network and literally have your car be an edge device that you can use for payments, securely. So I think it's just one example of how we can continue to extend the reach and create all sorts of new experiences for consumers.
How do these innovations that Mastercard's doing, how does it differentiate you versus the competition and some of the peers?
So payments has always been, 20 years at Mastercard, an incredibly competitive marketplace. And I think we've always excelled by providing the best experiences, the most reliable processing around that, and making sure the offering was tailored for the environment that it wanted to be in. So there's lots of different methods that are out there, but I think for the core consumer experience, what we have is even with the existing payment products we have, we have things that people love, that work great, and now we can extend it to all kinds of new environments for them to be able to use it. So as digital experiences become more immersive, as more and more of our life becomes device-based, those all become extensions of the network we already have.
So we can compete by having value-added services that are based on data and experiences that no one has. We have consumer propositions like Zero Liability, settlement guarantees, dispute handling. The experiences and the extent around the brand really puts us in an incredibly unique and I'd say, advantaged position from what we see in other alternatives.
You mentioned the virtualization of everything.
Yeah.
You know, whether that's tokenization or cards or other technologies. Maybe you could just help us discuss how Mastercard is driving these capabilities, explain how it works, the technology, and discuss why it's important to your strategy.
It is essential to the strategy. For those who've been following this for a long time, I think it was, check with Devin, 2012, where the foundation of our investor day is every device is a commerce device. At the time, it was even questions of we have, you know, you're card-based. And we had to explain that no one benefits more than Mastercard from being free to the tyranny of the physical world, because the whole point of the card was just to get payments onto the network. So we built the technologies, we pioneered the technologies for the industry to allow us to basically instantly issue a secure Mastercard credential into any environment or any device. And that has been profoundly transformative.
So what it allows us to do is everything from the mobile phone you had to, you know, your account with a streaming service, to your Mercedes-Benz, make that just a natural part and extension to the network. And what we've seen by moving to the virtualized credentials, because a lot of times I would get the questions: "Well, what compromises do you have to make to move into the digital world?" Digital is better. It's digital supremacy. So I see the plastic card as sort of a backward compatibility device to an infrastructure that already we have in place that people love. But if you look at what we're now doing with tokenization, we have a platform we call the MDES, Mastercard's Digital Enablement Services. We launched it a decade ago. Now, over 25% of our transactions are tokenized.
It's growing at 50% a year. When a transaction is tokenized, we see 3%-5% greater authorization rates. That's you as a consumer having a better experience. We see reductions in fraud. That's everyone doing better because the systems are safe to work with, and we can create and craft all of these new contextual commerce experiences like we've talked about. So when you have a tokenized card with, say, your streaming service, you don't have to worry about it being expired. If the card, for whatever reason, is replaced, we can automatically update it up there. If anything happens to the number that's stored there, it can't be used anywhere else, and we can simply reissue it.
So what we've done is taken the opportunity to move fully into the virtualization and the digital world to create a much better consumer experience, greater authorization rates, and reduce fraud around it. And that virtualization applies now to all sorts of new segments, like commercial. If you think about B2B payments, you have another flow which is never carded. It's open for us, but the type of capabilities we have in our network apply amazingly. So one great example is the online travel agencies. They need to, in a B2B context, pay hotels, pay airlines, pay travel services and cards, and things like that. They all take cards.
So the ability for them to automate their systems by pushing card numbers out and using that to work with the existing payment mechanisms, with all the capabilities of the Mastercard network, again, it's transformative, and it's been a great growth business for us. So the whole movement of procurement to payment, you know, we can provide an essential component to that. So companies like Oracle and Coupa and SAP are all embedding in Mastercard's virtual card solutions because there's a segment of the payments that are poorly served today that we can do a great job addressing. And the unlock is when it's not a card anymore, when it's a fully virtualized Mastercard credential that we can embed anywhere. So those are just two examples of that.
But whenever people say, "Well, what about this new environment?" I always go back to our foundational premise, right? Every device is a commerce device to make and receive payments, and that's really what the network's enabling.
So I got a surprise question.
All right, cool.
How are you thinking about Gen AI? You've never heard that question?
Yeah, although I was asked that a while back, and I did say, I think speaking for technologists everywhere, one of the greatest things about Gen AI is we got to stop talking so much about the metaverse. So that was just a win. That was just a win right there. No, Gen AI is an incredibly powerful new technique that we're looking to take advantage of. But I also think, and this is important, maybe even for your investment thesis, we need to put it in context because AI is white hot again. You know, Mastercard has been working on this for well over a decade, and it has already transformed our business. It is already the source of incredible value-added services that we can put in the network.
So let me talk a little bit about how we're using AI in general, and then how I see generative AI really extending and augmenting what we're doing. Yeah. Because if you're a Mastercard, one of the big advantages we had, and, yeah, I think we've used AI better than most, but we also had an opportunity that wasn't available to many, is we had an incredibly strong signal coming in. Was it fraud, right? We adjudicate all the time. We know that from our training set. Was the transaction authorized? Should it have been authorized? So we had incredibly rich data, and we had really big data. We did 143 billion transactions last year. Our transactions have been doubling every five years. Transactions, as a percentage, are flowing into our network. So we had a really big training set.
So when we applied AI, we could create value-added services for our customers, things like what we call Safety Net. It's a circuit breaker we run in the network. When we see a fraud pattern, we know we can stop those transactions before a human agent can even get a page. Like, and we do that, protect everyone in the network. Our Decision Intelligence, when we moved from the old rules-based systems to AI-based systems, we had tremendous, you know, increase in the precision by which we could spot fraud. I think it was about a three X. But we had a six X benefit in removing false positives. That's more business going through the network because we could tell, "Hey, this actually is good," even though we might have thought it was dodgy in the past. Better experience for consumers, more revenue for all participants.
We've doubled down on that, what we're calling our Auth Optimizer, where if a transaction is declined, we can say, "Hey, here's when you can retry it. Here's another way you could do it," and we get a point or two more authorization rate. Again, more business, better for everyone from using the sophisticated AI techniques. So we won Forbes Magazine Innovation of the Year in 2019 for our decision management platform, for the AI engines we're running. And we've reached the point today where we have over a trillion parameters running in our in-memory decision management system that we can apply and use AI for. And we're constantly using new techniques. So we have about 14 techniques we're using today. We'll run in what we call silent mode, where we validate, are the new techniques doing better than what we had before?
And then we can cut over to that. We're constantly maintaining those things. So AI has been massively transformative for us and allowed us to generate a tremendous amount of value. Now, in Gen AI, new set of techniques, we have what we're now calling Decision Intelligence, what I talked about earlier, DI Pro, which is using recurrent neural networks and other generative techniques to just like a ChatGPT works by guessing the next word and therefore generating incredibly compelling and sophisticated text. We can now look at anonymized payment histories of any actor in the network and guess what's happening next in a way we couldn't. So it's taking the attention, spreading it wider, and getting much more fidelity. So again, we've seen a 20% increase in fraud detection on top of all the amazing things we had already been able to do with AI.
So that's where I was saying when we talked a little bit about it earlier, you know, it's not a new game, but it's an amazing new club to have in the bag to apply to those use cases where it makes best. The other thing it does is because of the natural language interface and the ability to work with unstructured information, it allows us to apply it to new areas of the business. So I'm obsessed with being a place where everyone can do their best work. It's one of the greatest parts of my job. So we've talked about workbenches. What's a knowledge worker's workbench? What's a sales workbench? What's a software development engineer's workbench? What's a data scientist's workbench? And how we can use these new AI and generative AI tools to make them more effective, to make them more efficient.
And that whole human-computer interaction and the augmentation, which goes all the way back to when I saw spell check in the late eighties and thought it was the most transformative thing. This whole idea of this computer assistance for being able to do better work, we're seeing huge, huge, huge potentials there. We're also using it for the product set. One example, we have something called Shopping Muse, which can look at your behaviors, talk with you about what you like, and then do massive research to give you recommendations for things you might want to do. You couldn't really do that before at scale. You'd have to have a personal shopper. But now these type of things to help aid commerce, to help our merchants sell more, to apply the technology in that sector makes a whole lot of sense.
And then there's last one I don't think people are talking about as much as you would. I think you'll see a lot of benefit across a lot of sectors with this, and it's combining AI techniques. So generative AI, we just won a CIO 100 award for this one. I was really happy for that. So combining things like the ability to handle unstructured information and then bring it back into your workflows. So we did one where we get all kinds of requests in from our customers through a variety of channels. Getting that to the person most qualified to solve that problem. It's always tough, and if you route it to the wrong place, the customer gets delayed, as you got to figure out where it really should go.
And, like any longstanding company, I've been at Mastercard 20 years. Name changes. People call things different things in different regions. So, what we were able to do is use, again, generative AI techniques to look at all the different ways tickets had come in and where it was actually finally resolved. And, we're now able to get, with one hop from the request, directly to the person who was qualified to do it. And, we had 10 people - about 11 people, I think it was, who all they did was route and try to get things to the right place. We've now shifted them, so we're all working on the quality side, saying, "We got it to the right place. How well did we service that customer?" So, those are the kind of compounding benefits I think you'll see.
So, already transformed our business, already driving a tremendous amount of value, and I think these new techniques are going to give us all sorts of capabilities to extend that.
Yeah, I'm sure we could probably talk the full time-
I would love to.
... on gen AI. I do want to move on and ask about Mastercard increasing the percentage of total transactions that you switch on the network, and I know that's really important focus of the organization.
Yeah.
Maybe can you help us? What has helped you guys move those switch transactions?
Yeah, that's... I'm glad you said. That's something that's been transformative for our business, which I actually think might be a little bit underappreciated.
Yeah, I think so too, actually.
So if you look back five, six years, we switched about half of our branded transactions, which has been great for us. Now, we're almost two-thirds of our branded transactions or over two-thirds. I'm looking at that, over two-thirds of our branded transactions we're switching through the network, and that comes from a lot of places. We've opened up countries like Mexico and Japan, big economies where we didn't switch before. In a lot of markets where there's domestic players in switching, those transactions have actually come to our network. Why? Well, maybe because a lot of the investments I talked about earlier, where the things we do on the backbone, the speed, the reliability, as payments become more and more critical in environments around the world. What we just talked about with AI, right?
I heard the phrase, "AI works opposite of Moore's Law." You know, every incremental improvement you want, you have to double your data, you have to double your compute on that one. So our ability to scale, to see across the entire network, and then apply that to every transaction that flows through, makes the network that much more valuable. You asked earlier about competition. That's incredibly unique capability. So what it does for us is we now see more and more transactions in real time being switched through the network. And we've generated all kinds of new value-added services we can apply to that, whether it was the tokenization, the virtualization we talked about earlier, the AI-based fraud scores, the safety nets to prevent runaway fraud.
So what you've seen in the numbers over the last five, six years, is around the world, people bringing more and more transactions to the Mastercard network. And as those transactions are in-network, our ability to apply more and more value-added services to them.
Is there still room to go to switch even more to-
Oh, absolutely!
Two-thirds-
Yeah, yeah.
Nine-tenths, or?
Yeah, so it grows in a couple of ways, right? One of which is all the new segments and sectors that we're going after. So it extends us into new ways. You know, opening up transit, going back to where I started at Mastercard, is you know, billions of taps now all coming to the network, and that's great for us. We're in economies which are red-hot economies. So a lot of things happening in Latin America. Like I said, the cash that's still there in Italy or in Germany or Japan are all wide-open opportunities for us. And by having those transactions on the network, we can also do a much better job ensuring the consumer experience using that, which again, differentiates our products in and of itself.
So I think you'll continue to see more shift coming to the network. You'll continue to see, I talked earlier about our direct services access. So even if we're not switching the transactions, we can apply value to other transactions, which would be another growth driver for us. So on us or private label transactions now become in scope for our services. So we'll be touching more and more transactions around the world, and that's a great business driver.
Yeah. Another kinda highly debated topic is account-to-account payments-
Yeah
... and bank payments. You know, I'd love to get your perspective on some of the threats and opportunities of that, both in the U.S. and internationally?
Yeah. Well, first thing I'd say, particularly on A2A, is we all know, like one of the hardest things to do is take something that works incredibly well and replace it with something which appears substantially similar, right? So I start with what we talked about earlier. We have 100 million acceptance points around the world. We're providing zero liability for consumers and payment guarantees for settlement. The investments we've made in the scale and the services and the AI, it's a really unique environment we have, where in many, many sectors, we are the best way to make the payments. But A2A is also something we know incredibly well. VocaLink, for example, is a Mastercard portfolio company. I'm on the board of the VocaLink UK.
The U.K. has had one of the most advanced A2A faster payment systems in the world since 2008. It's also one of our biggest and most vibrant card markets, so if I want to move my investment account to my transactional account in near real time across institutions, and I use an A2A connection for that, that's great. When the money's in my transactional account, and I want to get on the tube in London, you know, we have 180 milliseconds to make sure that gate opens, and if it doesn't, there's going to be a riot. You know, that is an incredible application of our network, so I think you look at the sectors where things are best, and you see more of a combinatory effect. The other thing is, we work with governments around this.
There's twelve economies. I think about 30% of A2A transactions in the world are running on Mastercard technologies... so we can apply value-added services to that. Some are AI skills and capability, working with 10 banks in the UK for financial crimes on the A2A systems, by aggregating the data and applying Mastercard expertise for us. So we see real business opportunities, and everything we know about the art and discipline of running a high-scale national critical infrastructure payment network can apply here as a business opportunity for Mastercard. Then, if you look at what's going on, really interesting stuff with, for example, Pix in Brazil or UPI in India. You know, a lot of that is going after financial inclusion, which helps us, because as more people come into the system, we are the aspirational product.
They will want to move into credit. They will come even more firmly into the formal economy. So our long-standing, "How do we bring more people in? How do we encourage sustainable economic development?" That helps us. So Brazil is still a very, very solid market for us. And Pix, in some ways, can even be a little bit of a tailwind for that. So getting more people in the economy is getting more merchants transacting electronically. In some ways creates opportunities for us that we couldn't do by ourselves. So I think you'll see a continued modernization of these 30-, 40-year-old archaic ACHs. I think most of that wave has come through. And then we'll figure out what's the best way to do which transactions on behalf of consumers and how we can add value to that.
Got it. You've obviously touched on and mentioned value-added services a couple times.
Yeah.
So I just want to ask you, directly, just that being a big part of the business and one of the fastest-growing businesses in at Mastercard and helping accelerate growth, how do you think about tech and data assets required to maintain that strong growth rate in value-added services and your leadership position there?
Yeah, and thank you for asking about the tech side, because I think that is where we have such structural advantage for that. So I when I think about VAS, it's also kind of like a big topic, what's a value-added service? So I really break it down in two ways. We have all the things which are around our transaction and all the things we do during our transaction. We talked a little bit about that when you asked me about the switching side of the business. So when I look at... and I'll start with what we can do during the transaction, because it's a huge scale driver for us, where we can create a new capability, and it can apply to over three billion accounts globally, right?
That multiplier effect, that scale, the ability to take value and have it impact the whole network or the transactions we're seeing across the network. So the fraud scoring, the authorization approval rates to help people drive their businesses, the tokenization, virtualization, those are all value-added services that we can apply in real time to our transactions. One of the biggest investments I'm making is how do we get to, you know, event-based streaming environments, where we can do all of this in real time during the transaction and interact with our customers and other business partners in new and more flexible ways? So a lot during the transaction. When you think about around the transaction, before we have gateway services, we have all sorts of ways to make it easier for people to access and bring transactions to the network.
I talked about Cloud Edge and things of that nature and then after the transaction, the data is such an incredible asset. And I will say, when we talk a lot about AI, I think we have an advantage in how we've always emphasized responsible AI, right? AI is data. Data is AI. So the data usage bill of rights we've published out to consumers, the work we're doing with governments and regulators around the world to have sane and safe policies for how we use it, to let consumers know we only use it to make the service better. We use it for your benefit and again, this isn't new to us. This isn't fashionable.
We was fortunate enough to be one of the founding members of Harvard Kennedy School Council for Ethical Use of AI in 2017, I think is when we kicked that off, so it's been an essential part of how we do it and how we handle the services, and doing it right, that asset that the data is, to train AIs, to generate new services, to provide analytical capabilities, to help merchants with their marketing, to have more compelling loyalty services for our consumers, all of those come from after the transaction, and then there's adjacencies, things like Open Banking and digital identity and working with those type of things, which again, enhance the whole ecosystem, so that's when we talk about value-added services, and I think it was 19% growth last quarter that we were driving through that. Those are all the opportunities.
But the thing I think is missed the most is the flywheel. We touched on it when we talked about why are more and more transactions being switched by Mastercard. We switch more transactions, we get into new flows, we get into new segments, it generates more data. We can apply services to those transactions. We can use that data to create new and compelling things to make our customers more successful. And in doing so, that generates more transactions for Mastercard. So if you think about how the play works, more transactions gives us more data, which means we can have the most sophisticated services in the world, and our customers can be more successful, and we can apply it to new segments. And as we do that, it grows more payment processing on the network, which strengthens the services again.
So that's really where it's not you have payments and services. It's that flywheel and that interaction, which I think becomes so powerful.
I wanted to get your perspective on the future. When you think about some of the-
Just the future?
Yeah.
Okay.
Just the simple future. No, but when you think about other key technology trends that you guys might be exploring-
Mm-hmm.
You know, what, what might have an impact at Mastercard and the broader ecosystem, some of the things that you guys are looking at and developing?
...In my job, we are constantly looking at what are new technologies that can help us do our job better? And it's never technology for technology's sake. I think that's a mistake, which are fallacies often are. You know, GenAI, how can I use it? No, what is a really, really hard problem which will generate a tremendous amount of value that this can let you solve in a way you never could before? So we do a lot looking at how do we lower our cost, how do we increase our resiliency, how do we have that global network fabric that we can have incredibly fast, unbreakable payments around the world? And we're bringing a lot of technology and a lot of investment to bear for that, and a lot of things which are happening with, you know, streaming techniques, with event-based architectures, all of that helps.
But we do that every day to say: How can we be the best at what we do? There's also new opportunities which are opening up. So not only can we do that, how do we apply the new segments? How do we expand into new markets, leveraging capabilities that weren't there before? So everything from passkeys, transforming biometric authentication and e-commerce to create better user experiences and open up even more places where we can use cards. We talked a lot about contextual commerce. How do we say it's digital environments? I made a metaverse joke earlier, but digital environments are becoming progressively more compelling, progressively more immersive. We are living more and more of our lives device-based. How do we have the right technology to enable to do that? To stealing a Joe Strummer line, right? The future is unwritten.
So the other thing that we're spending a lot of time on is how to make sure we're architecting our networks and services for future, you know, government requirements or compliance requirements or optimizations for speed. The line we always say is, you know, instances getting a lot faster. So our ability to make sure that we have the fastest systems, continue to have the fastest systems that are out there, that we can clear transactions through the systems faster. We leverage that global fabric that we've talked about, where we can put incredible intelligence at the edge. I'm a big believer in edge computing, bringing it back to hubs that we can distribute either in third-party compute or in our own private cloud, and the ability to locate that around the world.
So we're really designing for an architectural flexibility, for an acceleration of everything, and constantly looking to what I call applied research. We'll never be a pure research shop, but we've already done work on, you know, quantum key distribution to make sure that we have the right techniques for potential future environments. We validate that in the environment, how we handle biometrics better, how we handle network resiliency and attestability better. All of those things just create incredible, really wonderful new opportunities for us.
Well, Ed, I think we're a lot smarter now after that discussion. Thanks so much. Thanks for being here.
So much more I'd love to talk about. Take care, and thank you for the time.