This is the third episode in our series on Payments Performance Optimization, where Drew Edmond talks to leaders across the ecosystem who are tackling the same stubborn question from different angles: How can merchants maximize transaction approval rates and reduce failed payments? Check out the previous episodes with Oban MacTavish at Spade and Brant Peterson at Worldpay for perspectives from other industry stakeholders.

In this episode, Rehman Baig, Chief Product Officer at FlexFactor joins Drew to talk about how FlexFactor is rethinking payments optimization through a Merchant of Record model. For most people in the industry, “Merchant of Record” is usually associated with global tax compliance and marketplaces, not optimization. But FlexFactor is using it to directly address failed payments and auth rate challenges in a way that merchants simply can’t do on their own.

This is a fascinating conversation for anyone working to solve performance issues at scale, especially if your team has ever looked at your decline rate and said, “We’ve tried everything. What now?”

 

 

Drew Edmond: Hey everybody. I’m Drew Edmond, an Associate Partner at Glenbrook and your host for this episode of Payments on Fire. This is the third episode in our series on payments performance optimization, where I talk to leaders across the ecosystem who are tackling the same stubborn question from different angles.

How can merchants maximize transaction approval rates and reduce failed payments? If you’ve been following the series, you know by now that optimizing payments is never as simple as flipping a switch. If it were, this series wouldn’t exist and neither would half the tools in the payments tech stack.

Approval rates are shaped by everything from transaction metadata to issuer bank preferences, to how a merchant routes, retries, and authenticates a given payment, and even with the sophisticated orchestration and retry strategies in place, many merchants still face maddening inconsistencies. A legitimate card might be declined for no clear reason only to be approved a second later on the same checkout.

Today’s conversation presents a very different approach, one that flips the model entirely by taking over the transaction itself. My guest today is Rehman Baig, Chief Product Officer at FlexFactor, a company that’s rethinking payments optimization through a merchant of record model. For most people in the industry, merchant of record is usually associated with global tax compliance in marketplaces, not optimization, but FlexFactor is using it to directly address failed payments and auth rate challenges in a way that merchants simply can’t do on their own.

In this episode, we’re going to explore what that looks like. We’ll talk about how Flex Factor navigates risk, issuer behavior, data control, and customer experience, all while operating, under their own MIDs on behalf of the merchant. We’ll also talk about why protecting the integrity of your merchant identity matters so much in today’s landscape of failovers, smart routing, and hyperaggressive retry logic.

This is a fascinating conversation for anyone working to solve performance issues at scale, especially if your team has ever looked at your decline rate and said, we’ve tried everything. What now?

Rehman, welcome to Payments on Fire.

Rehman Baig: Thank you for having me, Drew.

Drew Edmond: Of course. We always want to learn about your path into payments. Why you are where you are today, what got you into payments, how did you end up working for FlexFactor?

Rehman Baig: That’s a really great question. Funny enough, turn back the clock a little over a decade and I found myself working for eBay Inc, which is the holding company that owned eBay and PayPal at the time. And at the time, I was just more so interested in broader ecommerce and whatnot, not necessarily payment specifically.

As a brief aside, I think we need to have a fleet of children’s toys, EMV terminals, things like that to help get more kids into the ecosystem. So we’ll get there one day. Slowly but surely, as I worked with the folks at PayPal and as I started to help understand more about the implications of payments from a PayPal perspective, that slowly got me into payments, which was a really fascinating opportunity to learn the ins and outs from, even back then, a global provider of money movement and payment services and wallets for people everywhere, all across the globe.

Funny enough, from PayPal, I had the opportunity to come work with you, Drew. We both worked together at Square.

Drew Edmond: We buried the lead a little bit there. I did know you in the past.

Rehman Baig: That’s right, that’s right. I will say that we worked together on what I’ll call the OG payments team at Square. It was a fantastic opportunity to help build Square right around the time of the EMV deployment. That was great because we got to work on a lot of really hard problems that probably other parts of the world had solved, but for the US, we were working on a novel deployment of how do we help to make this thing happen? How do we help scale across the globe? How do we work on things like pin on glass? How do we work on things like moving to mid-market? A lot of fun things that we were able to make happen there.

Worked there for a bit, a couple years, went to a smaller company called Yapstone. Helped build their platform out there, which was fantastic. And then I spent four years at Adyen where I ended up leading global payment partnerships and product partnerships for Adyen. And it was an honor to be there for four years. And as I left in early ’24, I looked across the ecosystem and I was trying to figure out what was new, what was novel. I think the thing, if I think about my past couple of roles, particularly at Square and at Adyen, it was always on the cutting edge of like, how do you help move the industry forward?

And ’24 was a weird year for fintech, to say the least. And when I was introduced to the folks here at FlexFactor who are saying, Hey, we have a better approach to payment declines than anyone else. It’s fundamentally different. That was what got me interested and got me engaged. We’ll talk more about it in a second over the course of this podcast, but yeah, I joined as CPO and had a blast for just over a year now.

Drew Edmond: So this is our third podcast right now in this series that we’re doing on payment optimization for merchants. As you know, because you’re deep in the weeds of this every day, failed payments are a major problem for merchants. We’ve been coming at it from different angles through this conversation series with Oban McTavish from Spade. We talked to Brant Peterson at Worldpay.

There’s tactics that merchants can do to help address their problem, right, ensuring that they’re using the best practices in payments to optimize their approval rates. There’s things that the payment service providers can do that can help optimize their approval rates. But FlexFactor’s solution is kind of an approach that those entities can’t really do.

So I want you to tell us a little bit about your approach to solving this hard problem that doesn’t seem to ever go away.

Rehman Baig: Yeah. So first there’s a little bit of a view of how we see the world. And Drew, again, like we’ve been around the block for a bit so I’m sure you can empathize on this one, is that a lot of providers in payments, they’re incentivized to be good enough. But there’s a point of diminishing returns where it starts to, being able to push the bleeding edge of what’s possible, starts to become, it’s a cost of capital concern, it’s other concerns that are going on.

And so they get to a place where you can get to 90, 95%, good enough. And patch that up across any number of providers, any number of orchestrators, any number of whatever it may be, and you get to a place where you can get pretty good acceptance, pretty good coverage to the point that, again, acceptance is largely considered effectively a solved problem. But we view the world a little bit differently that says, Hey, there is a percentage of transactions that haven’t been optimized. There are a percentage of transactions that aren’t routable as they could be. There are processors out there that don’t have what I would consider basic technology that was available over a decade ago, and, it was novel back then, and now it’s kind of like just they haven’t had a chance to build it out, or they have multiple platforms through a series of acquisitions and now they need to simplify it and it hasn’t gotten around to delivering on that system.

So our thesis of the world is, I need to have effectively several PhD thesises on payments optimizations operating simultaneously. When it comes to the best of how to use alternative routes, how to use tokenization, how to use the payments technologies available to help make payments go through, I need to develop the best and the best of the best in terms of what that looks like.

And what that means is that I’m able to help understand how to make payments go through at the margins, if you will. And so the thesis here is that payment declines are not a solved problem. Payment declines are something that there are suboptimal solutions for. The average ecommerce merchant has between 10 to 15% decline rates, in our experience. We’re talking about the unfiltered raw truth of the matter. It’s about 10 to 15% decline rates is what we see there. And our belief of the world is that about a third of those are legitimate transactions that didn’t go through, have been insulted, and aren’t going to go through.

The idea there is FlexFactor believes that we can look at a transaction in real time and use our technology to assess what’s going on in that transaction and determine if it’s one of those third that we think that should be saved or not. If we believe that it should be saved, the idea there is we can step in, purchase that transaction from you, purchase that decline from you, and allow you to say to the customer, Great, your order went through. Your order’s on the way, here’s access to the service, or whatever else it may be along the way.

That is the view of our world where we can come in and say, We will step in and take on the risk of that decline, which will allow you as the merchant to, again, engage with your customer and fulfill that service and be able to optimize in a way that is not otherwise possible, simply because we’re able to take things like time in order to make a transaction go through versus others where you have milliseconds to make the transaction go through. And if it didn’t go through, we’re not going to take on that risk, which is a huge differentiation from what happens today.

Drew Edmond: Right. So the, at that moment in time, the merchant sees, if they use your solution, they see a success, the customer sees a success, and it’s up to FlexFactor to actually obtain a success at some point in the future?

Rehman Baig: That’s correct. And the idea there is that it all done with zero impact to the consumer.

So they’re not paying anything extra, there’s no penalties or add-ons or anything else like that along the way. And so it’s just meant to be the seamless, smooth experience that everyone wants for commerce in the first place.

Drew Edmond: And just to be really clear, to kind of reiterate this, because I think that merchants in the ecosystem that have looked at different tools and solutions out there to help them improve their approval rates, especially on the recurring side of the business, right, there’s the known solution providers that will kind of retry transactions through your MID, your own MID as the merchant. and ultimately try to capture, recapture that payment or capture that payment successfully.

You’re looking at it completely differently. You’re saying through the merchant’s eyes, they’re saying this transaction goes through, the customer says the transaction’s gone through. The merchant says the transaction’s gone through. Now it’s up to FlexFactor on the backend, if that transaction actually failed, to go and capture that payment on your own to actually capture the funds associated with it.

Is that true?

Rehman Baig: Correct. And so there’s two parts of the decision really that sometimes can be tricky. And so sometimes people try and pattern match it. They’re like, Oh, so you’re kind of like a dunning service, right, for subscriptions you just mentioned. Right. Or you’re offering something that is like orchestration or whatnot.

Not exactly right, because there’s two parts of it. One is, do we think this decline is actually a legitimate purchase attempt that should be saved? And there’s whole decisioning set that’s there that is, I would argue, completely new, completely novel, hasn’t really been seen in the payments world before.

And then there’s a second piece of it is like, Okay, we’ve purchased that decline. The decline is now ours. The product has been shipped, the service has been rendered, the economic transaction is complete. Cool. Now how do we help deploy the thesis that we had in the first place that allowed us to purchase the transaction, to make the transaction go through on the backend.

That second part may have elements of orchestration, but it’s all on our internal private networks as opposed to being able to leverage what the merchant has themselves. And that’s something that is, it creates a lot of opportunity for differentiation. And I think that’s something that, again, if you get enough people into a room that are payments professionals, and I’m sure the illustrious Payments on Fire alum can certainly attest to this, to some capacity, there’s all sorts of invisible arbitrage and optimization opportunities inside of payments that can be applied to very subtle things that are almost hard to detect.

But even something as simple as padding of a field and a payload can lead to different outcomes, which is silly, but it’s something that, again, if you’ve been around the industry long enough and you talk to enough people, they’re like, Yeah, that makes sense. We haven’t figured out how to like, capitalize on it, but it’s something that exists.

And that’s something that again, you can be better at. And that’s where we are.

Drew Edmond: I think you’re spot on. I think there’s so many elements of payment optimization where ultimately it shouldn’t have to happen, but it does. It does have to happen to make this work and to get the performance that you want as a merchant, especially at scale, right?

I think when you’re talking about these basis points or percentages that you’re able to improve upon and save, when you start to talk about some of these larger merchants, it’s really real money. Which is why there’s dedicated folks and businesses and different solutions out there dedicated to these.

Rehman Baig: Absolutely. The way I think about it is what we’re doing is almost MAG or MRC as a service, right. Being able to help bring the best of what’s possible, adding our own secret sauce on top of that, and make it fly. And so what we can bring is actually very similar to the Square thesis of, when we were back there together, was how do you bring a true world class, enterprise grade payment system to those that don’t have enterprise grade infrastructure? Which can be pretty compelling.

Drew Edmond: As we were talking months ago, when I was learning more about FlexFactor and why I wanted to bring you on, or one of the reasons why I wanted to bring you on, was I’ve been doing a lot of work with subscription merchant, recurring merchants, and so obviously very familiar with those recapture services for failed payments in that world.

With FlexFactor, though, it’s not just for recurring merchants, or maybe not even specifically for recurring merchants. It’s almost a better fit in some cases for transactions that are happening live, just ad hoc transactions. If you’re on a marketplace or on a DTC website, the customer is right there.

Like customer initiated transaction and it fails. Okay. Now what? You have a small window of time as a merchant to say, Okay, maybe I can fail over to my secondary PSP or maybe I’m super sophisticated, I’m going to change one field and see what happens, kind of thing. But you have milliseconds probably before that customer leaves. So-

Rehman Baig: That’s exactly right. Yep.

Drew Edmond: I think FlexFactor fits in better for this, where it’s like, time is of the essence. I need to make a snap decision on what happens here. So can you talk a little bit more about how you view that customer initiated transaction versus the recurring transaction?

Rehman Baig: Yeah, so, I look at it in terms of like, just to kind of spell out the three categories that we think about, right, is one is the customer initiated transaction. That’s the classic ecommerce purchase. I’m buying a pair of shoes or whatever and that’s going through, right?

There’s the subscription, right? Which we all know and understand very well, clocks in every month, every quarter, every year, whatever it is. Fine. And then in between, there’s the initial attempt at a subscription, which is technically a CIT, but flirts that line of like, is it a subscription or is it a purchase or whatnot.

The good news is that, as you said, that’s what we’re where we’re able to solve all three use cases and being able to help provide a response in milliseconds, that helps you say, Hey, this transaction, actually, think should be approved. We feel so confident in that, that we’re going to step in and purchase this from you and effectively say, Go ahead, render the service, provide the shoes, provide the goods, and be on your way.

Which is great from a customer experience perspective, because again, they don’t see that decline. They don’t feel that insult, they’re in their flow. But the moment that I like to emphasize is that, when that consumer clicked Buy or click Start Subscription or whatever else it may be, that was a moment that came over everything else. When all the hesitation, all the doubts, all the certainty. They said, You know what, this is the price point, this is the item, this is the everything that I want. I want it. Click. Capitalize on that moment.

And that’s what we believe in of being able to step in on that. It’s really cool and I think the idea is, we look at it both from a direct and an indirect perspective, because again, on a direct basis, if you take the numbers that I gave before as true, 10 to 15% e-comm decline rates, we save about a third of those. You can start to do the math and get to some pretty compelling numbers and saying, Okay, cool, 3, 4%, 5% had direct revenue impact in a positive way of being able to help make transactions go through.

And I think there’s the also the implications of lifetime value of that customer, right? Of like, Hey, that person had a great experience with you. It went through when otherwise wouldn’t have. And at that point in time, they can go and they can tell their friends and they can shop again and they can do all sorts of things. And it’s a longer term view of things, but we were confident that if you look at it from a long-term view, you will see the compounding results that lead to some double digit returns actually.

How do we help think about it in the immediate, which is the direct revenue impact? And then how do you look at the LTV impact as well?

Drew Edmond: Right, and so I would imagine then that retailers that have their goods or services are easily replicable by others, potentially, you know, a shoe salesman, whatever. You can, go to Nike.com. I can go to to Zappos. Different places where I can go to find this particular shoe, perhaps. You want them to have a great experience there because you want them to keep coming back and saying, this where I get it. So there’s that segment of retailers. What other segments have you found traction with in terms of how they’re viewing this opportunity?

Rehman Baig: It’s fascinating. I think it’s really across the entire ecosystem. We found merchants that are, small merchants, they get it instantly. They’re fighting for their lives. They’re fighting for growth. They get it instantly. And the number of people that I’ve spoken with, they’re like, Oh my God, this is a huge problem for me. Please help me solve it. And I think that’s great.

Mid-market gets it. Enterprise gets it as well. It’s just you almost have to take them on a journey for a second of saying, Hey, you’re running a hundred million dollar a year enterprise, that’s fantastic, but you’re leaving 12, $15 million on the table every year that could have been part of your top line equation. Let’s think about what that looks like.

And so, in terms of categories, we’ve had everything from education providers to clothing retailers to digital services to literally almost anything is valid here. I think the only consideration that we have here is like if you’re selling on the ultra luxury end of like 10, $15,000 a ticket, we have to be a little bit more precise on our, on what we end up, purchasing on that decline.

But I think for the most part, we’re relatively flexible to look across the ecosystem to anything that would be considered generally low risk from a card network perspective.

Drew Edmond: Sure, sure. Maybe the pushback happens as you go further up, maybe less the total volume process, but almost the sophistication curve of a merchant, right? It’s like, what more incremental revenue are you going to bring to me versus what me and my team of data analysts and operations people and engineers have been able to do.

Does that resonate or is it-

Rehman Baig: I think there’s usually a story of like, How much can you do? And there may be a little bit more of a sales process. For the most part, the idea is rather compelling, even at an enterprise level. It’s just there’s more people to get kind of the approval and buy-in from.

Because the idea there is like, don’t change your product, change your marketing, don’t change your checkout. Don’t change anything. We’ll help you make more money. That for the most part tends to overcome any ROI or other type of calculation because it truly is meant to be an invisible layer that helps more transactions go through than otherwise.

Drew Edmond: I would say, devil’s advocate to my own argument, would be that it’s incremental, right? It’s like, Hey, you’re doing what you’re doing, we’re going to try to do a little bit more you. What’s the harm here?

Rehman Baig: That’s the beauty of it. We’ve had merchants that are running on the largest, most sophisticated processors that claim the best or are generally soon to be the best. And still, we see 20 plus percent recovery rates. There’s a lot of opportunity there, simply because, again, we’re running things at a scale and at a perspective that is fundamentally different than what any individual merchant can do unless they have enormous resources, to deploy and even then we think we can still outperform a bit.

And then also from that perspective of we’re willing to take on risks. So inherently, everything in fintech has risk associated with it, right? And from this perspective, we’re willing to take on credit risk that says, Hey, we think that this transaction should go through. Whereas I don’t think the vast majority of, merchants haven’t really gotten there yet. There’s a concept of a tab at a bar, of course, or whatever else it may be. Sure. But I think from an ecommerce perspective, we haven’t gotten there yet. And maybe we will as the, AI continues to advance, who knows?

Drew Edmond: Where risk exists and where those can take on risk and manage that, that’s where the money is to be made in payments oftentimes, right? There’s a reason why an agnostic gateway that only routes transactions and isn’t in the flow of, in the chain of liability, is going to charge you a few cents per transaction and your payment processor is going to charge much more in, in most cases.

So, the way I think about it is you’re really introducing, and, well, you kind of referenced it earlier, right, the notion of orchestration. There’s a, at this point, hopefully a relatively clear model in people’s heads of what that means in terms of being able to connect to different providers and route, do your routing logic and failovers and identify which one is your primary PSP for this and that.

But you’re able to kind of introduce some benefits of orchestration without actually having to go through, I think, some of the complexity of setting it up and developing that where. Okay. I’m using PSP A. They’re performing, pretty well. But, as we know, maximizing approval rates requires having multiple routes because there is different behavior on those different routes.

Under the hood, you’re able to leverage that and, create a world where kind of everyone wins, right? I mean, the, the merchant wins. The consumer wins. You win. Everyone’s, everyone’s happy

Rehman Baig: That’s the intention here. That’s exactly right. I think the whole point is that, yeah, most merchants, most e-commerce merchants of scale that I imagine are probably listeners on your podcast have at least one redundant option, right? Backup option. Some will have the second one. And I think the idea is how do you figure out which one should be in line in a given transaction if the first processor fails? Sometimes you just have a random order. Sometimes you have some intelligence behind there. There’s a lot of options in that you can think through there.

But to have 10 or 15 or 20 different options is nonsensical, unless, there’s plenty of merchants, unfortunately, that have Heads of Payment that have been the recipient of acquisition upon acquisition upon acquisition. There’s 50 to 100 payment processors that are there, but they’re not all masterfully planned together and orchestrated. They’re just kind of dealing with legacy contracts at that point.

And so the idea there is, you can get to really interesting optimizations that are known to be true, again, by practitioners in the field, but just it’s not always practical to go and do them. But as an example, surprise, surprise, if you send Chase cards over ChaseNet, under certain configurations, it tends to perform better than if you send it to other processors and whatnot.

And so there’s little things that you can do, that again, if you know enough, you can make it happen, but is a merchant going to go out of their way and say, Hey, we have X percent of our volume there, which, again, should be quite large given that they’re one of the larger issuers that’s out there. But is it worth it building an entire processor, negotiating a contract, maintaining and upkeeping that connection? There’s a lot of overhead associated with all of this in the grand scheme of things.

Drew Edmond: And I think we’re in a world right now where, unfortunately, I think a lot of teams have downsized. A lot of people are being asked to do more with less. And so that limits your capacity as a payments operations team to say, Okay, yeah, I’m just going to add a bunch more providers or things like that.

What other tools can I use that might be a little bit more lightweight than a full PSP integration to get the outcome that I’m being asked to get to from my leadership team.

Rehman Baig: Absolutely. Yep.

Drew Edmond: So I’m going to put my old payments operations hat on. Yeah. It’s pretty dusty. Getting a little dustier every day.

I think anyone that’s a merchant or thinks through the merchant lens is going to be curious in terms of, Okay, this sounds great. What does this actually mean though? In terms of what is it the experience for, especially the customer?

Rehman Baig: The merchant side matters, but we’re being a representative of the brand.

Drew Edmond: Absolutely. So leaning into this notion of merchant of record, which it seems like you’re leveraging probably, in this case, and correct me if I’m wrong, but talk about a little bit, like what does the customer see?

The payment actually failed, so maybe they’ll even notice that the payment didn’t go through on their statement to the merchant that they were expecting to interact with. What does it ultimately end up looking like for the customer experience in a happy path and a less happy path?

Rehman Baig: So the idea here is we have to first remember that as payments professionals, we obsess about payments and the mechanics and the fit and finish of payments way more than the average consumer does. When we step in as merchant record, there’s two parts of it.

One part of it is like, how do we help merchants plug into us easily. Ideally, one click, which is a separate conversation. We can talk about that later. And then there’s the consumer aspect of it, or what does that look and feel like from a consumer experience side of things?

So, on the consumer side of things, we step in as merchant of record, as you just mentioned. And the idea there is, we’re PCI certified, Level One, we have all the requisite environment and practices in place to ensure that we keep cardholder data secure. So we take that same credential that was provided in the first place, and then we’re able to then operate that and say, Cool, you provided your X, Y, Z card, we’re going to charge that X, Y, Z card for exactly the same amount that you would’ve charged them in the first place.

And so what shows up on their statement is the dynamic descriptor that has our prefix, star, whatever the merchant name was. And so as for simplicity, you can just say, it says FF star, merchant, or whatever it may be. And again, you and I know that FF star means something. To the average consumer is just a bunch of letters.

It’s exactly the same transaction. It’s exactly the same amount. The merchant name is recognized. We are using all of the latest and greatest technologies that are out there, like consumer clarity and order insights to provide detailed itemization level data to help recall transactions and things like that.

There’s a lot of pieces that are going on to deliver a high quality consumer experience that, I would argue, is above and beyond what most merchants are doing themselves. And so the idea there is you’re providing the best, going back a couple answers ago, the PhD level thesis of what should payments look and feel like from a consumer side. You’re trying to deliver on that experience so that it looks like, Hey, everything’s going through as it should normally. And the fact that we stepped in is just a minor speed bump or deviation on that path.

Drew Edmond: You’re telling me the average person doesn’t read the card network rules on the format required for statement descriptors for aggregators?

Come on. It’s great reading.

Rehman Baig: If only we created that children’s toy line of EMV terminals and gave them a Visa manual, they too would read it and know, understand it.

Drew Edmond: Take it to the beach this summer.

Rehman Baig: That’s exactly right. Yeah,

Drew Edmond: That all makes sense to me. What about things like refunds and chargebacks? What’s the experience like for a customer there?

Rehman Baig: Yeah. So, refunds, the intention there is, as merchant of record, we have all the obligations that go along with being merchant of record, right? That’s there. However, what our preference is, to the extent possible is that, the consumer has complete optionality. If they contact us, we will happily help them.

And if they need a refund, we’ll give it back to them, because we want to be as consumer forward, consumer friendly as possibly we can be. Even if that means we’re eating a loss, we think it’s the right thing to do, especially at how we’re operating today.

If you contact the original merchant, which we think is the preferred way, and again, for the most part, we’re trying to be the invisible enabler. We’re not trying to be the active one that’s involved unless we absolutely have to be. They still have full rights over the transaction. And so if we stepped in and we purchase that decline, they say, Yep, let’s issue your refund. And guess what? We can go ahead and unwind that purchase. That’s not a big deal from our side. We were completely at the discretion of what’s best for the consumer is how we operate from that perspective. So the idea there is consumer has full control, which is key.

From a disputes and chargebacks perspective that comes to our MID, right. And so the interesting piece there is that ultimately it is our MID, it is our infrastructure. And so we do monitor that extremely closely and it’s part of our underwriting diligence process whenever we onboard a new merchant because we have things to manage at the end of the day here.

And so this is something that, at the end of the day, we are able to provide, representment. We can share information and itemization details and whatever we can to help show that the transaction was legitimate. But that kind of goes through the normal flows that already exist today.

Drew Edmond: Got it. And then going back to just thinking about retry logic and the things that merchants can do. There’s a spectrum of how aggressive certain merchants may be, or the providers that they use may be, in terms of how often they’re retrying a particular transaction. I think there’s this general hypothesis in the world today that that can lead to, I won’t say a disintegration, but at least a reduction in how the issuing banks can view a particular merchant if they’re hammering their systems with retries.

And maybe ultimately that might lead to lower approval rates in the future if the bank tightens their tools and their fraud models appropriately to reflect that. How does your approach help with that?

Rehman Baig: We’ve had the conversations with Visa and Mastercard where we ensured that we’re not hammering the system and just trying random things and whatnot. When we purchase a transaction, we try and be very purposeful about effectively sending the right transaction at the right time to recover the funds that were involved in that transaction.

And so for the most part, we are really only looking at doing two maybe three attempts over the course of a couple of days to make this transaction go through. And so as a responsible actor in the ecosystem, I think that that tends to be much more palatable versus what some of the other folks may be doing.

The answer is not just blindly hammer. The answer is be precise about what route, what optionality, what needs to be in the payload, what information needs to be helpful, to make this transaction go through. And again, that requires an enormous amount of data science, and it requires just being able to be right more often than you’re not. That’s the key piece that we’re able to apply that allows us to be more confident when we actually end up purchasing. Because otherwise you’re just gambling at that point in time if you don’t actually know.

Drew Edmond: From the merchant perspective though, how it helps is you’re kind of taking that on yourself. It saves the merchant’s MID a little bit if they’re not having to do that on all their transactions. Maybe it results in their volume being a little bit cleaner and hypothetically could lead to better approval rates for them.

Rehman Baig: Yeah, that’s exactly right. So that’s the hidden catch 22 with all the existing, especially on the subscription retry sides, where, it’s great that they have the service, I know exactly why they exist, all the billing providers and enablers out there, awesome.

But they’re just signals. They’re saying, Hey, it’s Friday. You probably got paid today. Let’s try it again today. It’s Tuesday afternoon. Transactions happen to go through more often here. Try it again here. Guess who’s on the hook for that? The merchant with their merchant MIDs, paying their provider per auth, per attempt, per whatever.

And, yeah, at the end of the day, it does lead to a negative decline rate. Which again, is one of those things that if you’re in the industry enough, you know that this is true. But for the most part, most actually don’t realize that this is the case. That having a decline threshold, if you will, will further impact your acceptance in some capacity. It’s not just every transaction is isolated, completely unto itself.

Drew Edmond: Given what your model has to be, right, it has to be quite confident in your ability to ultimately capture that transaction, given that you’ve outlayed this capital for each transaction to gain your benefits from it.

Part of that is fraud, certainly. That’s the game we’re all playing these days as merchants is trying to stop those bad transactions and to allow as many good transactions to go through as possible. How do you strike that balance, between optimizing for approval rates and keeping fraud levels as low? Which I guess for you represents the acceptance of the transaction in the first place.

Rehman Baig: Yeah, it is being able to help understand the peculiarities of the transaction in a way that, just like any fraud provider would do, we’re able to look at things like device fingerprint or IP address or other factors that may be applying to a given transaction that make us think, Hmm, this got flagged for whatever reason. Is this something that we think is legitimate or is this something that we think truly, truly shouldn’t be approved at all?

What’s interesting is that again, if go back to the thesis that says, let’s say 10 to 15%, decline rates that are out there. Over half of that is legitimate, right? Like over half of that should be declined, full stop. Right? It’s fraud. It is someone trying to spend on a gift card that has a dollar on there that will never see another dollar again, ever again. There’s all these things that are in place that we need to be better at.

And so the good news is that we’re really confident those aren’t happening. It’s just for the other half that is up for consideration at that point in time, the more information we have, the more certainty we can build. So it’s almost like a second look or a second pass for any given transaction that goes through.

Drew Edmond: Yeah. We talk about the issuing banks obviously having a, maybe the most important role in the approval rate process, given their ability to ultimately decline or approve a transaction on their end. You’ve spent a lot of time on the acquiring side, right, the Squares and the Adyens, or on the merchant side. And thinking about it from that angle, have you learned anything from the data or from conversations that you have about just how issuers view this problem, about approval rates in general?

Rehman Baig: It’s something that ultimately they have conversations on with the card networks. This is something that’s part of the normal portfolio review that occurs, and there’s certain expectations of performance and the acceptance and whatnot that every bank goes through this ringer at some point in time.

I think that the question really ends up being is, which banks have the right level of investment and infrastructure in their issuer processor and in their issuing risk system that they have to be able to meaningfully understand how to move things back and forth to come to the right outcomes.

Of course every bank, their core competency is risk. That has to be their core competency. But in terms of how that applies in a cards world, oftentimes this ends up being outsourced, right? There’s a default set of rules that a lot of people were able to end up deploying, a default set of models that people end up deploying, that unless you’re a top 10, top 15 bank, you’re not really going to make that investment. And certainly if you’re a credit union, certainly if you’re a community bank, you’re basically just saying, listen, I focus on other areas of risk, not this for the most part. And so yeah, being able to help figure that out yields different outcomes.

You see silly things come out of the ecosystem where, when I was at Adyen, before Adyen moved to its own banking infrastructure and its own BIN sponsor, or, I don’t know. As a bank, it became a direct member of the Visa, Mastercard ecosystems.

At the time when we had two banks that were our sponsor banks for the BIN, you could take the exact same merchant, the exact same transaction, the exact same consumer. All the details are the same. It would fail on one, it would go through on the other. And obviously one off stuff happens. It’s not that big of a deal.

Drew Edmond: I mean, that’s wild, right? That’s just an acquiring BIN, right? That’s not even, literally nothing is different.

Rehman Baig: Exactly. It’s a BIN level perspective of how performance happens. And it’s again, if you, speak to enough folks at the majors, if you will, to the right people that know what’s going on, they will say, Yeah, we observe this. And why it happens, who knows? There’s all sorts of little reasons why that could happen there. But the idea there is there’s lots of opportunity. And so if there’s a belief that a transaction could go through, if only under the halo of a different bank, that’s something that’s interesting that could be explored and as there’s opportunity there.

But again, the vast majority of merchants, it’s not worth their time to go and optimize at this level. There’s massively diminishing returns as you go down this pipe and businesses have other priorities, and that’s okay. How do we help figure out how to enable that? And that’s where we come in.

Drew Edmond: Totally. One of the hottest phrases in payments these days is agentic, fill in the blank. Commerce, payments, whatever it might be. Some sort of AI. We’ve got to talk about AI at some point during this conversation or else we’re doing something wrong.

It’s still pretty early days, I would say, in terms of this notion of agentic commerce and agentic payments. And for the uninitiated, we’re talking about an AI agent that’s perhaps discovering what to buy, initiating a purchase, completing a purchase on the behalf of somebody.

As consumers potentially begin to delegate more decision making or more payment activity to these agents, it kind of changes the structural way that we think about how payments work, in terms of how they’re approved, how they’re optimized, the risk associated with that particular transaction, et cetera.

But I’m curious, your viewpoint on this world, how does it affect you? Is it something that you can leverage in some way? How do you think about this notion within the world of FlexFactor?

Rehman Baig: Yeah. I’m going to go out in a limb and be bold enough to state what I’ve stated publicly before is that I believe we are the world’s first agentic payment service for the benefit of a seller. I think there’s a lot of people out there that are trying to crack agentic commerce, agentic payments for the benefit of consumers, businesses as the buyer, or whatever else it may be.

But this is a world where we will effectively come in, look at a transaction in real time, make a snap decision, take that transaction on, and then complete it on our own regards. Regardless the merchant gets paid, whether or not we make the transaction go through on our side. That to me feels like agentic payments, because there’s an independent system that is deploying AI technology along with classic machine learning to be able to make these decisions on the fly and get you an answer in less than a second. I think that that is a compelling view of the world because again, everything is going to be about how do you supercharge AI from acceptance, from issuance, from everything in between.

And we’re big believers in that and we’re making a ton of investments to help make this happen and make this fly as well. So, we’re excited for that future. There’s a lot of nuances that people are trying to figure out. And I think that this is one of those important ones. It’s all new and exciting and fresh again, which is kind of cool.

Drew Edmond: Yeah, absolutely. All right, and then maybe our last question here as we wrap up. How do you generally see kind of merchant payment optimization evolving over the coming years?

Given that you’ve really immersed yourself in that world now, I’m sure you talked to a lot of folks across the ecosystem that are looking at this from different angles. Given that I’m focused on this for this podcast series and trying to see all these different angles and perspectives.

What have you seen? What else is happening out there?

Rehman Baig: I think what Stripe and Adyen are claiming to do with AI optimization. I think there’s an element of marketing, I think there’s an element of truth along the way, but I do believe in the vision that they have out there that like, Hey, we can use AI to better optimize and ideally to even better request and facilitate information to help payments go through.

And I think that that’s going to be a key part of the conversation between merchants and processors or whatnot over the course of the coming years. But again, it’ll take time to come to fruition. It’ll take time to make sense of the data. It’ll take time to be so bold as to make the bet that says, Hey, if you can finally pipe through 3DS, not a trivial ask or anything else like that, this will help increase your approval rates under these circumstances or scenarios for your, American listeners that are here. Europe or other parts of the world, it’s a different story. I think these are the things that processors are going to be leaning towards, and that will naturally, merchants will kind of fall into that.

I think the question ends up being does the place of a consulting firm that historically has maybe done manual analysis for merchants and get to a place where like, Oh, well here’s the opportunity and here’s the cost savings, whatnot, move to a much more automated and AI driven process. And I think the answer is yes. It’s just a question of how soon, what does it look like, and then what are the results? What do the results mean at the end of the day?

Drew Edmond: I think that’s right. I think the technology, obviously, that we’re being exposed to these days and how rapidly it’s advancing is going to make a dramatic difference. We’re starting to see it in things like what Stripe and Adyen have talked about and what some of the other PSPs are doing as well.

I guess I want more folks thinking about this on the other side, the bank side to help them make better decisions, or what’s the network doing? The other pieces that are like really critical to having this work completely. Because I think if just one side does it, it’s like how it’s been historically. It’s the merchants and the PSPs trying to solve the problem, not the root cause as much.

Rehman Baig: As we all know, and with all respect to my, colleagues and friends, whatnot, that are on the banking side and network side, it’s a hard problem. There’s a lot of policies, a lot of governance. There’s a lot of regulations that they have to go work through and deal with.

And so, I respect the problem set and I think that the natural emergence and the rise of AI, even in business, is going to help these businesses kind of consider these decisions faster. Which is an interesting world that I don’t think people are fully prepared for yet.

I’ll share an interesting tidbit that I heard over the weekend talking to a friend, another part of the world. There are parts of the world where AI is a first class member of government advisory boards. It is literally as valued as much as a human in that process.

And if governments have been able to make that jump and leap, I think banks will be able to figure out how to make that jump and leap as well. It doesn’t have to be today. I don’t know. It’s something that is going to come over the course of the next couple of years.

Drew Edmond: It’ll happen. It’ll happen. Absolutely. Fantastic, Rehman. Well, this has been a great conversation. Thank you so much for joining me on this episode.

Rehman Baig: Thanks for having me. Appreciate it.

Drew Edmond: And to all of you listening, thank you for joining us. Until next time, keep up the good work and goodbye for now.