Payments Performance That Moves Revenue with Klas Bäck, CEO & Co-Founder of Pagos | Episode 517

Payments can be your second or third biggest cost line and one of your biggest levers for growth, yet most companies still treat payment performance like an afterthought until revenue dips. Greg Myers sits down with Klas Back,, CEO and Co-founder of Pagos, to unpack why enterprises keep “leaving money on the table” through avoidable declines, misconfigured vendors, outdated card network programs, and fragmented reporting that hides what is really happening.

We get practical about what payment optimization looks like when you operate globally: approval rates, authentication friction, dispute and chargeback signals, and the compounding impact of a bad first purchase experience. Klas explains why the hard part is often not strategy, it’s payment data. When information lives in silos across PSPs, acquirers, orchestration, fraud tools, and 3D Secure providers, teams spend weeks normalizing spreadsheets before they can even diagnose a problem. Pagos approaches this as a payments data platform, focused on aggregation, normalization, monitoring, and surfacing opportunities teams can act on.

From there, we dig into how AI changes payment operations. Klas shares how automation can shrink the manual workload, improve detection, and apply a huge knowledge base from card networks like Visa and Mastercard to real merchant data. We also explore emerging forces like agentic fraud and agentic commerce, plus why benchmarking and “time to detection” should be core KPIs for modern payments teams.

If you care about enterprise payments, payment analytics, and building a smarter payments stack with fewer resources, this conversation is for you. Subscribe, share the episode with a payments leader, and leave a review so more teams can find it.

Transcript

Welcome And Guest Introduction

SPEAKER_00 0:00

Welcome to the Leaders in Famous Podcast, where we talk to sea-level leaders from across the payments landscape and we'll be discussing the products and services that impact the payments space today, as well as trends and predictions for the future of payments. We will also hear stories from our guests about their journeys to the topic.

SPEAKER_01 0:18

Hello, everyone, and welcome to the Leaders and Payments Podcast. I'm your host, Greg Myers, and today's special guest is Klaus Bach, the CEO and co-founder of Pogos. So, Klaus, thank you for being here and welcome to the show.

SPEAKER_02 0:28

Oh, thank you so much for having me. I feel like I've made it being in this group of many, many, many famous people who came before me. So it's exciting.

SPEAKER_01 0:34

Yeah, great. I'm glad you're

Klas’s Path Through Payments

SPEAKER_01 0:36

here. So before we dive into the meat of the conversation, can you give us a quick snapshot of your personal background, maybe where you grew up, where you call home today, a few things like that?

SPEAKER_02 0:44

Yeah, sure, absolutely. So uh grew up in Sweden, had the opportunity to a couple of years after college work on a project in LA. It's about 25 years ago. I came out in January and realized people live in a warm climate. I'm in, so uh have been pretty much here ever since. Also the same amount of time I've been in payments, a bunch of different startups, pretty long journey with a Swedish company called NetGaro that got sold to Digriver 2007, worked a little bit at Digriver as a result of that. Then uh joined some friends at another small little startup called Braintree, raised a bunch of money, bought Venmo, eBay acquired us, and then spun PayPal out. Then we went with PayPal, PayPal went public and did a good job of keeping the Braintree and Venmo team together for almost six years, at least on the leadership team, as part of that journey, and then eventually left out and started Pogos together with my co-founder Albert, about five years ago coming up here.

SPEAKER_01 1:37

So

Why Pagos Exists

SPEAKER_01 1:38

okay. Well, tell us why you started Pagos.

SPEAKER_02 1:41

I think of the journey of Braintree being one of the larger PSPs, especially from a global perspective. We were local acquirer in 49 countries in Europe and North and South America, as well as Asia. Every single one of our customers had the same problem. They were leaving money at the table when it comes to payment optimization, meaning they could have done better, which would have translated to more sales and cost reduction and cost and payment, especially from those who have a lot of sales from an online perspective, can be very material, sometimes number two, number three, biggest line item in the whole company. So something to stay on top of. And I think the flows and the scenarios that you've got to follow to optimize that getting more and more complicated every year. So a lot of people are struggling with it. So it's not a single company that we have ever run into that didn't need help to do better or could you help to do better. And obviously, you think about it through failed transactions or things that you can optimize. But on the other end, you also have buyers trying to have good experiences from all these brands and they fail. And if that's the first-time buyer, they may never come back and they'd be very frustrated. So just a lot of opportunities to dig in and help companies optimize. And I think if to kind of throw a little bit deeper if into what we're actually doing, if you do that as a payment processor, for most companies it's too hard to pull everything out that you're doing to optimize moving further. So a core part for us, we don't process payments, we're a data company that helps enterprises optimize all their payment infrastructure so they sell more reduce costs and have their customers.

SPEAKER_01 3:09

Okay. So sounds like the problem was you know, a lot of times a data problem you helped optimize their payments, and that's what you set out to solve. How has that changed over time or has it?

SPEAKER_02 3:19

It really hasn't, but in a way, we started POGUS because we knew that was a need pretty much across every single one of our customers, and wasn't just Braintree or PayPal customers, any PSP's customers. So companies just needed help and their infrastructure getting more complicated, they go global faster, the cost aspects that we already talked about getting more and more complicated. So accelerating field, and Albert and I we chatted for a while about someone needed to fix this, and then eventually someone became us, and we started doing and building something about it. So Pogas was born.

unknown 3:51

Okay.

AI Changes Payment Operations

SPEAKER_01 3:52

Well, let's dive into one of the hottest topics today, which is AI. And obviously, I'm sure it has affected what you do. So, how is it changing the way that payment teams operate today?

SPEAKER_02 4:03

I think there are obviously a wide end of the spectrum. I think there are many companies we in particular work with enterprise companies that aren't necessarily always the fastest, but I think we're clearly quickly moving into a world where teams are being asked to do a lot more with less resources. I think in payments, even the really good companies when it comes to payments performance and payment optimization and having metrics that they are proud of, did a lot of work that was manual or time consuming or labor-intensive. So I think very quickly people are changing over. Like, hey, we can do this with less people and we can be more effective. How they execute on that could be something that Pagos helps them with as well, which I'm sure we get into. So that's one component. Depending on how you slice it, I think obviously a lot of companies, depending a little bit on what they're selling, are struggling with what I think of as agentic fraud. Like the fraudsters are getting more and more and more sophisticated. It's very hard to manage that. So it's very real also in how do I actually become better on making sure I don't lose my shirt, uh, so to say. So that's another one. And then obviously the payment industry as such are very busy talking about what agentic commerce will do to transaction processing around payments. I think for now, maybe the real action there is happening more on the discovery and how people find rather than the payments, but obviously a lot of discussions that will lead to a lot of consequences that people are trying to think through ahead of the time as much as possible. So putting it all that together, you're gonna see a lot of change. Are you gonna see people having to build excellence in payments with less resources? I think they will be able to do so and have a much more sophisticated operation than maybe their peers from five, ten years ago who build really big teams around payment operations and payment optimization.

SPEAKER_01 5:47

So if our industry is getting more complex and you know, these bigger companies, you know, they're global businesses, probably most of them taking payments from across the world, and it's harder today, probably than ever, and challenging and complex. But you're saying they probably are doing it with less people. How do you reckon that?

SPEAKER_02 6:06

That's clearly a trend we're seeing, and we are obviously part of accelerating that. I think some of the things that needed to be done were just done by individuals, whether that is optimizing and reconciling and understanding how they can do better around their costs. They have multiple vendors, different data that lives in different silos and different formats. Often, if companies were doing, because a lot of people actually don't do anything about it, so they don't even optimize around it. But if they were trying, let's say, optimize their costs, a big chunk of what people were doing, just pulling together all that data. So basically being a manual data engineer trying to get the different formats and standards in one format, normalize it, harmonize it, see, all right, where are their opportunities, then understanding all the complexity that's being rolled out in terms of the programs from the card schemes and what can they do, what are the tools they should use in order to do better and reduce their cost. Just the basics of understanding the lay of the land of your own operation took a lot of manual work. That could be completely removed by working with someone like Pagas who can just put that data together for you. And in our case, we automatically use EDINS to serve as opportunities where you can do better. Because I think one of the reasons why so many companies haven't optimized their operations in round payments as well as they could. Just the knowledge base that you need to understand all the different programs that are changing several times a year. The card schemes are rolling out new rules and policies that you need to follow. If you are in multiple verticals in terms of buyers or sales channels, you have different payment methods, whether it's something basic like PayPal and Venmo or Apple Pay or Google, or you're just going internationally faster, you have all these different payment methods. Just so much that you need to know and being able to be on top of in order to just begin building business cases for how do I optimizing that? So getting that to happen automatically, pretty key for companies to be able to be much more effective, even though they have less resources and less people in the team, versus maybe if they were similar position five to ten years earlier.

Metrics Monitoring And Faster Detection

SPEAKER_01 8:09

Okay. So what should the payments teams that we're talking about be spending less time on today? And then conversely, where should they be spending more time?

SPEAKER_02 8:17

Two things. One is obviously doing all the things that I just said, because agents and software infrastructure can do that so much better. And there is a reason why we haven't run into a single company. And we work with many of the largest companies in the world that are selling and billing online. That's a reason why we haven't. There's always a reason that we have been able to help them do better than what they did before. So that's one. Another thing that I think is very important, and very few actually do well. You do got to monitor all your key metrics. So in general, you probably need to monitor more metrics than you currently are doing. And doing so is could have historically been a lot of work. We cannot do remove that, but then you need to monitor all this increasing number of metrics. And things are changing in payments all the time. I think people think often that some of the metrics are a flat line, that's not the case, so you've got to be on top of it, whether it's a slow degradation because some program was rolled out and you didn't follow it, so it starts building over time, or something got pushed out, the code got pushed out part of your system or part of your vendor system that negatively impact the cost or approval rate, whatever it may be. All those things is something that people need to monitor and need to be alerted, and they need to be on top of how do I identify what that was and do something about it.

SPEAKER_01 9:29

Okay. So if these teams are getting smaller and they've got to perform better, how should the performance of the team actually be measured? How do they know they're doing the right things and being successful?

SPEAKER_02 9:40

Yeah, so I think the big step is actually establishing metrics and then track yourself over time. I think a lot of people don't do that, so they have metrics, but they don't necessarily perform a review of me versus myself. So that's a step one. The second step is really taking it, all right, I think I'm doing a pretty good job. Is that true or not? So you versus the market. And that can then lead to benchmarking that could be sliced more generic view, or you can be very specific for a subscription merchant in this country with an AUV on average order value of a certain amount selling to predominantly younger people. All right, let's find others that we benchmark again. So that will be the other big step that you really need to do, and that you need to monitor that it stays that over time. So something happened from a performance perspective? Did just you have that problem, or is it just one of your processors that have that problem, or is it the everyone in the industry, which is sort of a macroeconomic thing? All three are bad, but two of them you can do something about, and the sooner you do something about it, the less it's gonna cost you if you keep it simple.

SPEAKER_01 10:41

Okay. So when something happens that they're tracking, why is time to detection such an important KPI?

SPEAKER_02 10:48

Yeah, it's an excellent question. And maybe even for us, me and Albert, when we started pogress, we felt there was a need to help people do better. But even with that kind of viewpoint, we'd have been surprised how many how rare it is that people actually monitor in the first place. So catching things have a tendency to not happen for the majority. So when they do find them something, they may not have a system in place where they actively monitor. So it's more like something happened, it means you're financially off versus your budget. Finance starts asking questions, and that's when you start realizing something went wrong a quarter ago, two quarters, three quarters ago. So that I think is really a big part of it. When they do monitor, I think companies typically do that by people, and people don't work on the night or the weekend, and maybe don't have the tooling to actually accurate monitor, or they go travel, so therefore they didn't monitor that certain week. I think overall, many things get unnoticed unless they become really big. But all the other things that people should catch doesn't get catched until much later, to exactly your point. And when people do catch that, they celebrate, like, oh, we find this problem, we fixed it. Meanwhile, it's been going on for a really long time. And if you're honest with the assessment, you should have caught this a lot sooner, and you would have less versus your budget if you did. So time to identify and do something about those issues should be a key KPI. And extremely rare that companies think like that. They would say, if I met monitor things, I actually celebrate the fact that I catch it, not how long it's been going on or how long it took me to do something about it. So it's an excellent point. We fully agree.

SPEAKER_01 12:25

Okay. So you talked about the AI agents. What can they catch that a payments person professional might miss?

SPEAKER_02 12:32

There are a couple of ways of slicing that. One is just the sheer amount of knowledge that it takes in all the scenarios, visa MasterCard, push out bulletins with thousands of pages several times a year. Here are all the changes, you need to follow them. If you follow them, you will have perform better, cost or approval rate, whatever it may be. If you don't, you're gonna pay a lot more. So just knowing how all those things play out, that could be a scheme that just being launched. So, how do we forecast the impact and are preparing for what you need to do to stay compliant or best performing, if you can put it that way? Or it could be something historical. Companies start as a small group and then they grow and more and more revenue, and then it's like, hey, you really have an opportunity to do, I don't know, something super basic, debit routing in the US. What is debit routing? How do I do it? So it's just surfacing those opportunities and tools, which there are so many of that you can use in order to drive better performance, which is more successful transaction going through, or doing something about your cost, or better manage your fraud, or better managing uh chargebacks disputes. And if you widen the surface, there's also a lot of other things that you really should measure. All your vendors are they performing as well as they could, or have they done something misconfigured you, or something changed when they pushed out software that you needed to be on top of, or are you using, I don't know, two-factor authentication in Europe, three secure, or in Japan, or whatever market you're selling. Does all the buyers that are trying to buy successfully manage to go through that? If not, you have a conversion problem. So it's similar to the payment processing issue or your fraud rules things classes on the website for the first time. You actually think I'm a good, a bad guy, but I'm a good guy. And the system blocks me from buying. I'm not going to be happy about that transaction. Even though that was not real even a real payment transaction, didn't even go to the payment system. So it didn't get sent to the acquirer or PSP, and it didn't reach my issuer, so the issuer couldn't even make a decision if we should approve or decline this transaction.

SPEAKER_01 14:29

Okay. Well, we talked about how agents can detect things and obviously do it probably better than humans. But how can AI help teams understand the sheer size or the cause or the urgency of the payments issue?

SPEAKER_02 14:43

Yeah, I mean, so there's the detection part. It's also applying the knowledge base. What is this? Is this good or bad? Okay, it's bad. What is the reason behind that it's bad? And what can I do about it? Or how do I do something better? So I think it's all those steps where you can take all the payment knowledge and combine it with your data that needs to be very accurate and reliable, which a lot of companies have struggling with historically. So combining really good, clean, aggregated, normalized data with all the payment expertise in the world, and then surface ability to both identify what should be done, how it compares to the market, if we keep it simple, benchmarking, and how to best execute that. I think all that is incredibly complicated to do, even for the really large companies with huge payments team, because there may be people that know a lot about any individual topic in the broad ecosystem of payments, but not every single one of the individuals in the company know that. So depending on which part of the system they are touching, it's not as easy to do that when it's humans that are in the loop and need to both detect and know what to do about it. And answer the question, okay, we did something. Is that looking good or bad compared to what it should be or could be?

SPEAKER_01 15:54

Okay.

Clean Data And No Code Integrations

SPEAKER_01 15:55

Well, you mentioned something there that I think is something that we skip a lot when we talk about AI is the data part. Like, how are you helping these companies make sure they have clean data? They're probably getting it from hundreds of sources. So how do you manage that?

SPEAKER_02 16:08

Yeah, so a big part of what we're doing is really getting people's different payments data, if I can call it that broadly. So data from their own operation, data from their vendors, which could be an acquirer or several or PSPs, depending on terminology we use, but it could also be a payment orchestrator, a fraud provider, a 3D secure provider. So a lot of different things around. Different silos, different formats, different structures. Some provide a lot of details and context, some provide very little. They're using different terminology. So big part for us, how we help, start helping our customers, taking all those different data formats and structures and silos and aggregate and normalize it continuously in real time, a feed that you now can start building. Okay, where are the opportunities? Have something changed? What should I do? Something about it. So really data aggregation normalization is the key, and then put everything else on top. And I should say, many of our customers get started using our no-code integration. So we already built integrations to all the major PSPs out there. So there's no even any work to get started. Press of a button, all the data starts streaming into our platform, good things start to happen, and they can build from there with our platform. So no work to get started, big financial impact, and you will do better than if you didn't use it. It's been quite a good sales pitch. So as a result of that, we've seen tremendous amount of extremely even the largest companies that you can find from a payments team perspective, signing up and finding opportunities where we can help them be better, which makes them all look so much better towards their senior management.

SPEAKER_01 17:36

Okay.

Pagos Business Model Explained

SPEAKER_01 17:37

Well, what is the and it's a very generic question, but I think it's important to understand what's your business model? Because I know there's some data companies out there that charge based on how much money they can save someone, you know. So just maybe walk through that.

SPEAKER_02 17:49

Yeah, so maybe a little bit to frame that question first. What we're doing is basically three things. One, we provide sort of the data aggregation normalization and help companies monitor it and surface opportunities to do better. So kind of the core platform. We also provide API toolings around some of the tools that you need to optimize. It's hard to use if they're sitting on your payment processor side. Why? Because if you have more than one, they have likely built different ways of how they manage any individual tool that you can use to do better. Some don't even support some of the tooling. So you need to move those toolings on your end. So those are our API-based tooling that we basically help you enrich your data. And then thirdly, we provide what I think of as market context services, such as benchmarking or decisioning tools. So one, two, three, if you keep it simple. As a result of that, our pricing model is a fixed fee per month, which is a function of how much data you have, how many data sources and complexity. The API tooling is transaction-based, and some of the market context are based on results. We're going to help you recover more transactions successfully. As a function of that, we're going to take a cut of the outcome. If you don't use us, you're going to get nothing. If you use us, you're going to make more revenue, let's say, and we'll take a cut of that. So all three combined, or one or the other, depending on how many of the services you are using.

SPEAKER_01 19:07

Okay. Yeah, I'm glad you talked through that because that makes a lot

Team Skills Layoffs And Key Takeaway

SPEAKER_01 19:10

of sense. So let's go back to the payments teams for one more question, because you know, there's so much in the news about layoffs, and I mean, obviously in our industry and in the tech industry across the board, a lot of times it gets AI gets blamed. But how do you think AI changes the size and the structure and the skill set of payment teams going forward?

SPEAKER_02 19:30

Yeah, I mean, personally, I've done a lot of the big companies that have been laying off people the last two, three years. I'm not so sure it's really about AI yet, as much as it's maybe a knee-jerk reaction, a very delayed one, to the overspend and the metrics that people were tracking towards as a function of COVID, like, wow, my curve is going like this, I need to invest, and then kind of that faded out, and people are back to a typically higher level than where they were before, but still nowhere near what their investment thesis was saying a couple of years ago. So is it really AI yet, or is it just a function of we have too many people for where we are? Of course, individual teams and individual processes and developers and whatever have you are incredibly much more productive than they were in the past. So I think the really good teams are managing, okay, how can we do so much better with the people that we have? But not everyone is in that fortunate situation. They're just like, hey, we have too many people for our actual performance in general, maybe less so. So I think those two converge, but it seems to be an easy excuse to just blame that we need to trim our headcount because we overinvested a few years ago, to be honest. But obviously, we can expect that that this would change. And I think some of the AI tooling, as we talked about, hey, you used to have a data team of eight people that were doing a bunch of manual job work for 30% of that time. So you can go down to five people and they can use a third-party tool and perform better than when they had eight people. So some of that will play out, and of course, some companies are very fast on executing change of technology and vendors, and others are incredibly slow to go through that journey. And of course, you know, maybe the engineering side changes faster than some of the opsides. So you also have within the persona where some of those changes are happening. And for that, I I think we're still in the very, very early innings of that changes, but it's certainly it's accelerating. But it does give people the opportunities to have the same team do much better. And I think we should remember that as well. It's not just use it as an excuse to not get in legal trouble when they have to do layoffs because they have too much cost versus revenue.

SPEAKER_01 21:35

Right, right. Okay. Well, one final question. What's the one thing that you would want listeners that are listening today or watching the show? What would you want them to take away from this conversation? Obviously, we've covered a lot about the company, a lot about AI, a lot about the future of payment teams. So, what's the one takeaway you'd like people to take away from the show?

SPEAKER_02 21:54

If you keep it simple, payments matters a lot for the C-suite. So you do need to have a payments team and you do need to have metrics that you hold those teams accountable for. Those teams need to do two things. They need to optimize, and there is a lot of opportunities to do better. So, of course, you would want your team to contribute to more revenue, happier customers, and reduction of cost. And the second thing they need to do, things change all the time. They do need to monitor and they need to be fast on catching the things that inevitably would change, or you're going to have a pretty significant, nasty surprise compared to your budget. Every company on the enterprise space needs to move in that direction, and we can help them do so.

SPEAKER_01 22:34

Okay. Well, Klaus, I think that's a great way to wrap up the show. So thank you so much for being here. I know your time is very valuable. So thank you again for being on the show today.

SPEAKER_02 22:41

Thank you so much for having me. It's great to be in the company of so many other great payments people who have been here before. So hopefully we can contribute to that wider group.

SPEAKER_01 22:50

Absolutely. And to all you listeners out there, I thank you for your time as well.

SPEAKER_00 22:53

And until the next story.com, where you can subscribe to the show and where you'll find our photo. If you enjoyed listening, please share on your social channels as well.