Subscribe
Share
Share
Embed
Our flagship series will propel you to the forefront of the global ecommerce revolution. From analyses of breaking current events to the intricacies of navigating cross-border sales and regulations, Business over Borders entertains and informs any audience who wants to learn more about how international ecommerce works.
Leo Tucker:
Welcome to Business Over Borders. I'm your host, Leo Tucker. This podcast is brought to you by the Reach Network where we help you become a global payments expert. I'm joined today by Lohrasp Seify, our Senior VP of Technology at Reach. Welcome to the show.
Lohrasp Seify:
Thanks, Leo. Great to be here.
Leo Tucker:
So tell us a little about what you do at Reach.
Lohrasp Seify:
As you know, Reach is a financial technology company. So there's two different parts, two different big major parts in the company. There's all the things that we do in the financial sector, but on top of that, we use technology to try and make financial processes automated, cost less, more efficient, and at the end of the day, cheaper. So anything that has to do in that technology arm is what I try and help out with.
Leo Tucker:
Is there a project you're particularly excited about?
Lohrasp Seify:
What am I not excited about? I think at any given point in time, there's a few things that are happening that either no one else is doing, and we're trying to experiment with it and try and make it as good as possible, perform as well as possible. I would say generally speaking, anything that we do with projects that optimize auth rates or have a new foreign exchange, trading patterns, something like that are the things that I'm most excited about because usually the industry stagnates when it comes to those.
Leo Tucker:
So can you dig a little bit deeper into authorization rates? I mean, as a credit card holder, I'm aware that generally my credit card works. Is that not the case for everybody and for businesses who collect credit cards?
Lohrasp Seify:
When you swipe your card, there are a lot of things that happen between the money being charged to your credit card or being charged to you taken out of your bank account. However, however you've set it up to the receiver of the money getting the money. There's a whole bunch of players, there's a whole bunch of states that a transaction needs to go through.
Leo Tucker:
Can you walk me through kind of a quick "explain it like I'm five" end to end of that money flow?
Lohrasp Seify:
So I think the easiest way to think about it is people would normally think of it as if you're a merchant and I'm a consumer, when I go and swipe my card to buy a coffee from you, they think about it as if I am the one who's paying you where, in fact, it's your bank requesting money from my bank. So that's what we call the issuing bank and the acquiring bank. The acquiring bank is the bank of the merchant or the seller trying to acquire the funds. The issuing bank is the bank that is extending the credit or issuing a card for an individual. So technically speaking, it's not you and me doing anything with one another. It has nothing to do with Lohrasp doing anything with Leo. It has to do with your bank going and charging my bank. Now in that path of acquiring, we call that acquiring, you go and acquire the funds in that path of acquiring, depending on the size of the transaction, which geography you're in, the path you might take might be different.
The players might be different. There might be multiple payment service providers that do it. It might be gateway services that are there. It might go through fraud systems that there's a whole bunch of gates in between. So when we talk about the authorization rate, basically an authorized transaction is when the issuing bank authorizes for the funds to be released to the acquiring bank. And then just very overarching wide brush terms. When an authorization fails, it's because a piece in between has stopped the transaction, has said, no, I don't want to transact anymore. And like I was saying, that could be because of a company in between. It could be because of the bank themselves saying, Hey, no, I'm not going to authorize this. It could be because of a fraud system saying, Hey, the transaction pattern looks like fraudulent, so I'm not going to let this happen. But the transaction might fail for one reason or another.
Leo Tucker:
So one reason might be I take my trip down to Mexico and go to the local taco shop and I get declined. First I call my bank and say, Hey, what's going on? And say, I didn't know you were in Mexico. And so is that an example?
Lohrasp Seify:
That's a very good example actually, that the banks have the fraud systems themselves. Funny story that actually happened to my partner. She had a $1 charge somewhere in the US but just because the timestamp of that transaction was very close to a transaction that happened in Calgary and it was a retail store, a physical store, they automatically locked her card after a $1 charge. So the fraud systems that are being developed are actually in production right now, are very good at it, at catching fraudulent things like that. When someone that's not the owner of the card that's using the card, we generally just call out of all the transactions that have happened, how many of them approved and succeeded. We call that the authorization rate. So that's what the authorization rate is.
Leo Tucker:
So what tools do you have? Do you use employ to make a difference in authorization rates? It seems like almost something that the banks are in charge of. How do we get in there and move the needle one way or the other?
Lohrasp Seify:
The path between the acquiring and the issuing bank could be a smooth one, but it could be a rocky one as well. So if the merchant's in Calgary and the consumer's in Calgary and both of them work with one of the big banks in Canada, things will be pretty smooth. The issue in the acquiring bank are under the same umbrella. Things are happening locally. The cardholder is there.
Leo Tucker:
Yeah, my credit card generally works. Yeah.
Lohrasp Seify:
So you go to a shop, you swipe it, it works as long as you have enough funds, it's not as simple when now we're separating the operating geographies between the banks, and you could think about it as countries, but it's not really country by country. But just think of it as the domain of operation of a bank doesn't really overlap too much with the domain of operation of other banks. Maybe their standards are not the same. Maybe they don't know each other well enough. Maybe one of 'em is a new bank and the other one isn't. So when I was talking about the path, depending on the players in between and what acquiring banks are used in that path, you can reduce or increase the friction of a transaction. So one of the benefits of Reach, when we talk about international transactions, is exactly that. If two banks play nice with one another and all the players in between, suddenly auth rates go up. There's less reasons to decline things.
Leo Tucker:
So if the authorization rate is between the acquiring and the issuing bank, we try and make sure that the acquiring and issuing bank are friends in a way that they understand each other. They speak the same language.
Lohrasp Seify:
Exactly. Okay. So an example of that is if we want to process a transaction, let's say the consumer's in the UK and the retail store is in the US. So the consumer's bank is in the uk. The store's bank is in the US to take the money from the consumer and give it to the retailer. We try and pair up the banks that are most friends with one another. So in our particular case, we have banks both in UK and we have banks in the US. So if we make it so that the consumer's bank only has to talk to our bank in UK, things are local and they're friends, and then we ourselves as the merchant can transfer the money somehow to our American bank pretty simply as well.
Leo Tucker:
We're all friends in this world.
Lohrasp Seify:
And we're all friends. So we turned something from drawing one line from one bank to the other one, and crossing a border to two banks in the same locality, talking to one another, two banks in the same locality talking to one another. And then the big circle in between is Reach. So we handle all the complications there.
Leo Tucker:
That's really interesting. But that must cause a peculiar, that must cause an issue with how many banks you're dealing with, how many friends do you have and whose friends and who are not friends. How do you keep track of that? That must get extremely complicated.
Lohrasp Seify:
Yeah, that's an extremely good question. It is extremely complicated for a specific payment method and a specific currency, we might have tens of different paths that a transaction could take. What we do is we try to collect KPIs, collect data on what the authorization rates given any of the paths that we could choose are going to be. So let's just make it simple. Let's say we have two paths. We put a specific payment method and a specific currency through both paths. We measure the authorization rates in real time, and then when we collect enough data that we decide, Hey, this is statistically significant. We make a decision to send this type of payment to this type of path. So in a very data-driven way, we make decisions that benefit us because we're now authorizing more, benefits the retail store because the store’s authorizing more and benefits the consumer. So the consumer is not angry pulling out their hair, Hey, why is my credit card getting declined?
Leo Tucker:
And you must have tons of data gathered on this over a long period of time. Is that kind of the trick is what to do with all that data and match everything? What is that project like?
Lohrasp Seify:
Yeah, yeah. Like you were saying, it's a sea of data. I guess in the modern world, everything is about big data on how you handle big data and making sure you're asking the right questions and making the right decisions using it, not just collecting it and just analyzing it just for fun. So we've been building our tool. We've been iterating over it and constantly making it better for the last two and a half years, and it's still not done. It's a lengthy process. And look, you have two options. You could either take all the legalese, all the things that are in the contract, all the intricacies of how the banks talk to one another, what geographies they're going to, all of which is probably published in public information. And you can have probably a legal going through every single one to figure out which banks play well with one another, what their charges are, and what's the path of least resistance, which in my mind, it seems almost impossible. Or you can take this data-driven approach where you send an impulse, you measure the reaction. The reaction being what's the authorization rate? And kind of like a scientist, I guess. You test different hypotheses and you see what works, what doesn't. I wouldn't claim that I'm an expert when it comes to issuing and acquiring banks. I wouldn't claim that I even know why some banks are friends and why some of them aren't, because it'll take a lifetime to understand.
Leo Tucker:
But if you've got enough data in your set, I assume you don't really need to know necessarily the proof is in the pudding, if you will.
Lohrasp Seify:
Yeah. So we basically go for the outcome rather than the cause. The outcome that we're trying to get is for them to have the highest auth rates. Do the consumers care why the banks don't talk nice to each other, or which acquiring bank it was that was contacting their issuing bank? No, no. We try to make the life of consumers simpler, then go for the outcome, and we don't have the overhead of, I don't know, reading contracts. If you ever read a Visa or MasterCard contract in a specific geography, you don't want to read a second one.
Leo Tucker:
No, thank you. I'm not interested. Appreciate it though.
Lohrasp Seify:
You're welcome. Anytime. Anytime. That's how we do it. Going through a slew of data, try to make sure things are statistically relevant, making sure that the things we're analyzing are actionable and they're not just to quench our curiosity and actually have a measurable impact.
Leo Tucker:
That makes me wonder if there's been occasions where the assumption of what would work, like taking our earlier example of which banks might be friends, had an assumption, ran that through, collected the data afterwards and were completely wrong. Have there been any surprises in that domain?
Lohrasp Seify:
More than a few surprises, actually. I guess I'm not surprised that we were surprised because one of the first things you learn is you don't know anything, right?
Leo Tucker:
Don't be surprised.
Lohrasp Seify:
So we again, approach the problem of authorization rates with the mindset that we don't know anything. Obviously, the hypotheses that we come up with are rooted in some sort of expertise, some sort of proven fact, but since we're not understanding every layer in absolute depth and understand it 100%, we expect there to be surprises. We even built systems to catch those surprises. We call them outliers. So we built systems that catch the outliers to make sure that we're not making wrong claims, we're not going down the wrong path at the beginning. I'll just generally talk about the evolution of this and what are the surprises that we had.
Leo Tucker:
I'd love to hear that.
Lohrasp Seify:
We thought if just the issuing bank and the acquiring bank are basically in the same locality, no matter what issuing bank and acquiring bank, there might be our auth rates would be high, and why wouldn't that be the case? I just explained that should be the case. But then it turns out that if you have a list of 50 banks in a locality to pick from and another 50 banks in the other locality to pick from, there are pairs of banks that just like we were jokingly saying, don't like one another, so they will not approve. They won't approve transactions that might be coming from that one particular bank, which surprised all of us, which now meant, okay, not only do we have to care about the locality and some of the other features that define where the bank operates in, but there's sometimes specific banks that don't play well, and we need to avoid matching up those two banks with one another, which is something again that we do. I wouldn't say we look for the highest within those 50 banks, for example, we still abstracted away to locality, but if out of those 50, one of them is a no-no with another one, we make sure that that never happens. So there's no unexplainable, "Hey, why did my card get declined?" Especially because there's thousands of reasons a card can get declined. There's so many error codes for them.
Leo Tucker:
With making friends of banks, pairing them up, how is this offering provided to your customers? How do we make that available?
Lohrasp Seify:
It just is available. That's the beauty of it. I think, again, like I was saying, trying to abstract away some of the complexity and making sure that we're the ones that handling with all the cross border banks are not in the same geography complications. Consumers just get to use the benefit. So if you're checking out with reach behind the scenes automatically, we're doing everything that we can to make sure that your card's getting authorized. If you're a retailer we're doing the same type of work to make sure you're capturing as much of the volume. So basically anyone that wants to buy your product and they go through the trouble of putting in the credit card number and pressing submit or okay or whatever it goes through, and they don't have to, I don't know, go look for another card that's not declined, or maybe that's the only card that they have, et cetera, et cetera. That's the beauty of it. Since everything's happening behind the scenes, there's no extra complications for anyone. No one needs to understand the complexity of issuing and acquiring bank world. No one has to understand the complexity of the different currencies of the different payment methods, We just handle all that.
Leo Tucker:
That's great. It sounds like it's baked in.
Lohrasp Seify:
It's baked in.
Leo Tucker:
That sounds like no information I want to learn on my own. Thank you. Appreciate you doing that for us.
Lohrasp Seify:
You're welcome. You're welcome.
Leo Tucker:
Loh, thanks for joining us today. I appreciated chatting with you, and maybe we'll see you on the show again next time. And until then, thanks for joining us.