In payments, as it is in life, a simple question is often wrapped in complexity. The complexity of the payments acceptance landscape, with an increasing number of stakeholders and technology providers, changing regulations and rules, and Issuers implementing new authentication and authorization mechanisms – any one or a combination of these could lead to major and minor impacts to the operational and financial performance of a business.

And while this has been the case for several decades, we find ourselves in a macroeconomic environment that requires us to find operational efficiencies, which places the payments team under increasing pressure to maximize the number of successful payments, lower the cost of accepting payments, and support customer acquisition to drive incremental revenue.

Long story short, these “simple questions” are being asked more often and at higher organizational levels. Why are simple questions hard to answer? Yvette Bohanan talks with Klas Bäck, Founder and CEO of Pagos Solutions, and Glenbrook’s Drew Edmond.

Yvette Bohanan:

Welcome to Payments On Fire, a podcast from Glenbrook Partners about the payments industry, how it works, and trends and its evolution.

Hello, I’m Yvette Bohanan, a partner at Glenbrook and your host for Payments on Fire. Some of the most frustrating moments of my career occurred when I was asked a simple question and did not have a simple answer or embarrassingly any answer for that matter. My guess is that many of you listening to this podcast have had the same situation happen to you. You’ve been asked a simple question, something like, “So what’s causing the spike in card declines?” Or “How many new customers subscribed when we added that digital wallet?”

All but the most sophisticated merchants wrestle with monitoring and managing their payment operations and the data that comes from it. The data’s not always clear cut. It’s murky, nuanced. Sometimes it’s missing, and this can lead to those awkward conversations, whether you are the person asking questions or the one trying to answer them. So joining me today to explore why answering these simple questions can be confounding for payments professionals is Drew Edmond, one of our associate partners at Glenbrook and someone who helps our clients answer simple and tough questions all the time. Drew, welcome to the podcast.

Drew Edmond:

Thank you, Yvette. It’s great to be joining you today.

Yvette Bohanan:

So Drew, I have to ask before we bring on our guest here, do you have a particularly dreaded question that can be easy to ask and difficult to answer, maybe something you’ve recently helped someone with?

Drew Edmond:

Yeah, absolutely. We’ve been working with a lot of subscription merchants lately to help them reduce their involuntary churn, which of course is critical for increasing the LTV of a customer. I think given the macro-environment that we’re in today, companies seem really eager to optimize their payments metrics right now, and it’s really one of the key metrics we want to look at, is for all of your subscribers that are getting billed on some regular basis, of the transactions that fail on the first attempt during the billing cycle, how many were retried, how many attempts did it take to ultimately recapture that payment? Because we want to tie that information back to the retry logic that the company may have implemented or is looking to implement to recapture these potentially lost customers. And it can be really hard for some customers to understand which transactions are tied together, which was the first one, which is a retry, and having the granular details around decline reason codes and card types to enable the analysis that we want to perform.

Yvette Bohanan:

That sounds complicated. And so the simple question that comes from that is something like, so what’s going on with churn?

Drew Edmond:

Yeah. How effective is our retry strategy?

Yvette Bohanan:

How effective is the retry strategy.

Drew Edmond:

That’s the simple question. Then you have to ask all those questions.

Yvette Bohanan:

Exactly. Okay, that’s a good one. That’s a good one. Our guest today had this to say on the topic, and I’m quoting, “Payment processing can no longer be a checkbox for running your business well. It’s a critical part of long-term growth strategy and execution. We believe that demystifying payments via consolidated data and insight APIs can accelerate business growth and create long-term customer relationships or even enable new companies and new business models” end quote. My guess is he has a lot more to share on this topic with us and we’re looking forward to a fascinating conversation. So joining us for this episode is Klas Bäck, no stranger to tough and simple questions, and co-founder and CEO of Pagos Solutions. Klas, welcome to Payments on Fire.

Klas Bäck:

Thank you for having me on the show.

Yvette Bohanan:

Thank you for being here. It is always fabulous to see you. So we’re going to cover a lot of ground, but I want to start by asking you to share a bit about your career with our listeners. How did you get into payments and what led you to co-founding Pagos?

Klas Bäck:

Yeah, absolutely. I’ve been in payments a long time, almost as long of a time as I’ve known you, which is probably at this time longer than you almost want to admit. Once upon a time, I was an engineer in Sweden and I had the opportunity to get to California on a project, working more on the industrial world. But I got fascinated at the end of the nineties on e-commerce and one way of getting into e-commerce turned out to be a payment startup that I joined to lead the US operation. It was called Netgiro. We were really an early gateway that helped extremely large merchants connect to all the different payment providers they needed at the time when there were very few global partners from a PSP perspective. So they ended up having a lot of partners and we made that possible. We worked with large companies like PayPal, eBay, and Expedia. I believe Yvette one of your former employers when we first met long time ago.

Yvette Bohanan:

Yes. I recall a very, very cold February in Sweden, that was wonderful and absolutely positively the coldest I’ve been in my life.

Klas Bäck:

We should have been a better host and scheduled that for the summertime, I guess.

Yvette Bohanan:

Project deadline’s a project deadline, you can’t move that. That was beautiful.

Klas Bäck:

So from there we eventually sold that company to Digital River, Digital River used that to create their own payment process, which for a time was called Digital River World Payment, and Digital River I guess changed the focus a little bit and eventually spun it out to what I now believe it’s part of the broader Worldline organization. There’ve been so many roll-ups in the broader payment world, it’s sometimes hard to keep track of it.

Then after that, I wanted to actually get out of payments, thought it was getting a little bit boring, got to do something different, not sure what. So I took some time off, floated around a bit, and I had a bunch of friends in payments. So I ended up helping people a little bit, consulting with some merchants, some marketplaces. And then eventually I found myself one day spending two days a week with a friend of mine who was one of the starters for Square or Block as they’re called nowadays, and three days with a friend of mine who just joined to become a CEO of a small startup in Chicago called Braintree.

And then that eventually led to Bill Ready, who was the CEO at the time, he convinced me to join full-time, which I did in early 2012 and lead payment strategy and international. We raised a bunch of money, $75 million, bought Venmo along the way and got some really crazy traction very quickly and PayPal figured that out and came in and gave us an offer we couldn’t refuse. $800 million in cash. Was less well known though. Obviously the Venmo brand still lives strong and are in the public eye, but Braintree was really used as a platform for PayPal to expand into becoming a PSP or a merchant acquirer, really on the enterprise side, PayPal maybe not have done as much, and they let us operate Braintree almost as an independent company. And things really took off also within PayPal.

And the teams stuck around there, at least in the leadership level for almost five and a half years until 2019 when people start moving in different directions. So probably some sort of record, but sometimes, and by the time we’re probably one of the largest, especially most global payment service providers or acquirer out there.

But that leads then to the second part of your question is why do we start Pagos? Because I think every single customer that we had, no matter what country and what vertical or big or small, had a combination of the same problems. They didn’t really have a good sense of their data around their payments operation, which led to everyone basically leaving money at the table. They could sell more, they could reduce their costs and they could use the data to do better in terms of here’s my benchmark, how can I improve it, what are my challenges? And just leverage that. And for many, that is almost an impossible task. It’s a combination of payments can be very complicated. Using the data can be very complicated because it isn’t very easy to use straight off, so you got to massage it yourself. It’s not readily available or it’s old.

So it really only gets pulled out when something catastrophic has happened. And then by that time, yes, you can understand what happened, but not necessarily prevent it from happening again, always. So we want to make that simple. And on that premises, how can we help companies be data-driven around payments? How can they use that data to get insight and how can they use that insight to take action to improve their performance? And here we are. So two years into the journey, the new startup journey with Pagos.

Yvette Bohanan:

Wow, thank you. That’s quite a career trajectory. I love how you like, “Well, I just went to California” and then-

Klas Bäck:

Well, it was sunshine and I grew up in Sweden, so I didn’t really need more than that. Wait, people live in places where there’s sunshine in January? I think I’ll sign up here.

Yvette Bohanan:

I think I need to check this out. So as you went through the journey, you met a lot of merchants, you worked with tons of merchants, you worked with people in companies where there was a formal payments organization, sometimes there wasn’t. The role of a quote-unquote “Payments professional” has changed a lot over the years. What have you observed? We’re talking about the data, and we’re going to get to data and metrics quite a bit here in a moment, but the importance regarding the ability to get payment data in broader decision-making for the organization. You alluded to that when you were saying people were leaving money out on the table and they sensed that, but it’s getting even bigger now than that. And what does having access to payment data mean for businesses? When you’re out there talking to merchants, what are you hearing about that?

Klas Bäck:

There are a couple of things that been driving this over the last decade. I would say broadly speaking, everyone’s environment’s becoming more and more complicated. You build for many companies, organizations that sell in more channels, maybe used to have a website, maybe you’d have the website then a store. They were often separated in the early days of internet. Now they’re not. You needed to deal with mobile, but on mobile, people didn’t buy the same way that they did on the website. And mobile isn’t one environment. If you’re a little bit bigger than at least, then it’s a mobile web scenario. It could be an Android environment, it could be an iOS environment. Different teams, different technology, different platforms, different data, which means it’s more complicated to have an understanding of what’s going on and how to make it better. And I would say companies build businesses faster that go global faster, add more payment methods and just add complexity from a business perspective.

And then some of the business models are more complicated. How do you build an Uber and make payments really baked into the experience is quite different than maybe the early days of internet. It’s like it doesn’t really matter. People are going to come here and buy anyway. So more pressure for the structure. If you run a marketplace, you need to do payout, suddenly you have to pay in and payout to manage. And some of the stuff in the middle in some places get regulated. That adds even more pressure on top of that. So I think there’s two scenarios. Either people discover like, “Hey, payments as a strategy is actually really important for the outcome on the business” or the complexity forcing you to trying to deal with the consequences of the complexity that you now operate as a company.

Yes, the importance and the challenges for people managing or owning payments is certainly a lot more complex for so many more companies now over the last few years.

Yvette Bohanan:

Drew, I imagine you see this a lot with our clients too.

Drew Edmond:

You see the spectrum in terms of maybe level of sophistication across merchants, level of knowledge and knowhow and even just pure resources, you may have someone that has a decent amount of knowledge but they don’t have a payments analyst. And then you go up the spectrum and you have large sophisticated merchants that have teams of analysts and folks with a lot of payments industry knowledge. So you have this wide variety of ability to ingest data. Are you taking the data actually into your own data warehouses? Are you relying on some payment service provider that has a cursory overview of your data? Can you actually do anything with the data?

And then do you know what to do with that data in the first place? Are you building dashboards? Are you measuring your own performance? Are you measuring the performance of your providers? Do you have the level of granularity and the quality of data that you need to do all those things? So the ideal world is that data is fundamental to you as a payments team or as a head of payments or whoever you might be. But the reality is that not everyone has the resources or the knowhow to get to the outcomes that they want to improve their payments environment.

Yvette Bohanan:

I would agree, and I think you’re both raising really good points and you’re bringing back a lot of memories from when I was in payments operations, but I’m not going to go down that rabbit hole. I am going to bring us back because here we are, we’re going to try to unpack this to understand why few merchants can really operate in a way that’s data-savvy about their payments data, given all the benefits that would be there if they were able to do so, if they had the data available and knew what to do.

So here’s what I would like to do with the two of you for this podcast. I have a few quote-unquote “Simple questions” that I’d like you to explain what happens when people try to get reliable, actionable answers to each question. And the goal here is really to unpack what’s happening under the hood a little bit so that there’s some common ground around the room or the Zoom call or whatever we’re using these days. When these conversations occur, when you have someone in the business asking the question, someone trying to answer it and it’s not a simple answer to that simple question. Maybe we’re building a little empathy here. So are you guys in for this? Are you good?

Drew Edmond:

Let’s do it.

Klas Bäck:

We are ready.

Yvette Bohanan:

Okay. So first question, we’re going to start with cards and I’m going to pick something pretty straightforward. What is the average order value for each form of payment a business accepts? So let me rephrase it in a slightly non-geeky way. A CEO shows up on a Zoom call and says, “We let people pay for a lot of different ways, which way brings us the highest sale, dollar value?” That seems pretty easy to answer. Why can that be hard to answer?

Klas Bäck:

The reality is at least for a little bit larger companies, payments gets very complicated very quickly as a structure of how payments are priced and get reported. I think for many, can you get the data in one spot? Already there the answer is no. So you may have to pull it together from several sources, but then when you then have the data, let’s say you look at your payment service provider’s report, you often, at least when it comes to card, have to do the roll-up of counting all that manually yourself before you even can look at the data. And that doesn’t really take into consideration it’s changing over time. So then you have to do even more math yourself to even answer something as basic as that.

But the CEO is likely not answering for a totality roll-up for one month or one quarter. The real magic in answering that question is maybe breaking it down. So is that different across products or countries or is it different on sales channels? Is it actually different across vendors and how can that be if that is the case? In reality, it’s a very straightforward question that just answering on top level is hard for almost all companies. But the real answer is then in the details, how can I break it down and how can I take in consideration what time does to that? And was it something we did that improved it that drove that payment method X actually is now higher average basket size or average sales number than it was last month, or was it an accident and it’s a higher good number or a bad number depending on the pricing model for that specific payment.

Yvette Bohanan:

Just reflecting on this for a second, it’s interesting how that question can arise. Because sometimes a question like that comes up because you’re looking at cost data first and classically there’s a lot of payment methods out there even just in the card space that will say, “We charge more to process our particular brand of card because we get a higher average ticket amount for you coming through the door,” something like that. Or if you add this particular payment method, you’re going to have incrementality, you’re going to sell more, whether it’s a private label or this or that. And you’re faced with the question, I might be all right paying to get that incrementality, but do we even know what our incrementality is here now? Do we even know which one of these today that we already have is giving us the highest ticket value? That starts the whole conversation.

Klas Bäck:

That’s exactly right. We have seen several people, actually, this is a real scenario or variation of that in real scenarios like, okay, we added this payment method, that was great, we got some new customers, but now it’s even more share of the total amount of payment that are being made on the website and this payment method it’s more expensive. So the reason people are asking, “Hey, you just drove up our cost for the whole company,” is that happening for the right reasons or is this actually a bad thing that this payment method is getting more and more successful? And if your margins are low or you’re looking at reducing your cost, then of course it becomes more sensitive in a C-suite level question in some companies.

Yvette Bohanan:

It’s interesting because it can depend a lot on how you’re monetizing that product. Is it a physical good at a high value or a ticket to begin with, or is it a digital good at a 99 cent or lower price point? How does that weigh into all of it? So we’re going to go back to, question two here is going to circle back on Drew’s example from earlier a little bit.

Subscription monetization strategies. They’re critical to many, many businesses these days. It seems like we can’t go a day without talking about recurring revenue and subscriptions around here. And there’s two sides of the coin, churn and incrementality. So here’s the question, which payment methods see the highest customer churn? And when we implemented payment X, did we get new incremental customers? It’s two questions, but who’s counting? It’s always peanut butter and jelly, they always come together.

Drew Edmond:

Hey, I can jump in on that one first. I think when it comes to the churn question, which payment methods are seeing the highest customer churn, first you want to make sure that you have a framework around how you’re measuring things and defining churn in the first place. I think that certainly you want to start at least with voluntary versus involuntary churn. Did someone choose to leave you versus is this a payments issue, a declined payment, and that’s leading to involuntary churn? So that we can dig into the payments problem there because if it is voluntary, maybe that’s a product issue, that’s something else that you need to deal with. But with involuntary churn, that’s a bad payments experience where for whatever reason that payment method declined.

Now when did they actually churn, if it fails the first time, have they churned? Are we waiting until the next billing cycle to determine if they’ve churned? Is it a year later? I think there needs to be some definition around when churn actually occurs because there’s a lot of tactics that you can take to recapture that customer along that timeline. I think making sure that you have an internal understanding of when you decide a customer has actually churned.

Yvette Bohanan:

And so many times people will show up with, “I’ve got the churn data” and they start looking at it and then everyone in the room or on the call goes, “Wait a minute, how are you counting this? What do you mean? That’s not how we think of churn over here in marketing” or wherever.

Klas Bäck:

I would even say eventually saying that it’s most likely that the answer is we are not looking at that the same way. So you don’t even know what the data mean because two people probably think about it differently. Like one of my favorite question is like, “Have you deduped the data?” It’s like, “No, what’s that?” And then it’s like, “Okay, we don’t even know what we are looking at anymore.”

Yvette Bohanan:

And then everyone goes, “We need to have another meeting. How much time does it take you to go dedupe and redefine and calibrate?” Drew, what else?

Drew Edmond:

And then speaking to the incrementality question, if we added a new payment method, did we gain new customers? You probably want to also make sure that it isn’t just your same customer switching payment methods as well. You don’t want to be double counting customers just because all of a sudden they switched to a new card or they switched to… Now they’re using Apple Pay and underlying it’s the same card at the end of the day. So I think there’s some work to be done to make sure you’re deduping on that side as well.

Klas Bäck:

I think when it comes to this question, I think the starting point is a little bit too like what we alluded to, are we actually looking at the right data? Is it easy to create the right data and can I from there establish a benchmark that I then can use to improve from? In many cases the answer to those two are no, or its too much work, or people don’t know how to do it, or they know how to do it, but it’s not easy from a practical perspective, the company where they work the data may be there somewhere, but they don’t have access to it as an example.

So getting that benchmark can be incredibly hard and it can also be wrong. And then you draw the wrong conclusions. Especially in subscription or retry, people may try to solve that then by brute forcing it, “I’m just going to retry everything three times, see what happens.” Like, okay, you’re retrying a lot of transactions that are an expired card. That probably is not going to help you. It may help you, but what’s the price for that from a cost perspective since there are consequences like that? So is that really helpful? Is it effective? Are you actually improving it?

And then a little bit back to what we talked about before, if there are no identifiers on the transactions, how is this related back to countries or platforms or vendors or product offerings and how does that come into play? A little bit like what Drew was saying, if you’re completely blind to all those factors, what do you actually see and therefore what can you actually conclude, if anything? And if you conclude it, is it even correct? So therefore what you’re trying to do about, it is that really going to change anything?

And I think if you can’t establish that, then you don’t know things like, “Okay, this was people who sign up March 2021 versus March 2022.” It’s like, “Oh, that’s right, in March 2022, we gave everyone 50% off and that’s why they were signing up.” And the behavior of that cohort is very different than the behavior of the court for March 2022. So even breaking it down on that and just getting to have more clarity on the details and the insight of what is really going on is very important. And we work with a lot of really, really big subscription merchants and I think many don’t have that clarity, which makes it very hard to be effective. And then people are doing things that are maybe just costing them money and not driving the result they’re looking for in the first place.

Drew Edmond:

I think you brought up something really important there. And when I think about the silos of information that can occur in an organization, when you talk about, “Oh, this cohort they were given a 50% discount, that’s going to have a significant impact on maybe the growth trajectory of customers in general.” And then if you’re just in a silo saying, “Oh, we added this payment method around that same time and look at this growth in this payment method,” it’s because they came in at the same time that you added this discount as well. So it’s not necessarily that payment method drove that growth, but these other growth activities in the business were doing so.

So I think that having that cohesive or holistic perspective on what’s happening across the business is helpful to understand how to actually read your own data and make sure that you’re not looking at it in a vacuum.

Klas Bäck:

I think that’s an excellent point. I think very few people actually do that, are all the different teams that are impacted actually using the same data looking at it the same way? Or they correlating activities in one end of the life cycle to how it impacts another end. So here, we’re talking about retracing subscription, but it could be how you set your fraud engine versus the outcome that in some industries are tied to chargebacks and very costly consequences of running that poorly.

Yvette Bohanan:

That end-to-end view in getting everyone on the same page and using the same report, is it undertaking in and of itself? It’s a big deal and there’s a lot of churn, no pun intended, created organizationally when people are trying to sort their numbers out with each other and they’re all carrying their own perspective. Getting something that everyone can believe in is a big, big goal here.

Okay, let’s see. How much does it cost to accept a payment method? Let’s dive into this one a little bit more, and here we’re looking for exact numbers. Pretend that you’re talking with the CFO, you’re talking with someone who wants to know down to basis point level. This is an essential question because payments can be one of the top three expenses for some businesses. They may not realize it is, but I’ve worked in organizations where the cost of payments was higher than payroll at some points, and that’s pretty significant. And these weren’t small companies.

So knowing the expense can actually give a company a big edge over the competition when it comes to their financial performance, because if you’re saving, it’s dropping dollars into your margin. There’s always a follow-up question to this one too. So the first part of it is how much does it cost to accept payment method X, Y, Z? And are we paying more or less than other companies in our space? That’s the competition question. Are we disadvantaged in the market because we don’t have a handle on this, or we’re paying our provider too much? It’s kind of nuanced, isn’t it?

Klas Bäck:

Yes, you can slice that discussion in many layers, and I would say card payments in particular, at least especially in the US, makes that even harder to answer than in other countries. So if people are thinking about it, one part of the cohort would think about it, how much am I actually paying my payment processor vendor? Even answering that question is not something that everyone can do easily, but if there are able to do that, is that really the biggest cost? No, typically it’s not even close to the biggest cost. It’s actually the consequences of your actions actually impacts the performance or the cost, I should say, even more. So how do you understand what your actions are and how do you link those to driving a better outcome from a cost perspective?

So what do I mean with that? If you break it down, the costs are in a couple of different categories. So you have interchange, you have assessment fees, you have processor fees that are tied to the specific transaction, but you have also a bunch of other fees. And maybe you have additional vendors too, that are impacting that. But then the data you send to them will have an impact on the cost performance as well. So you can do your retry strategy that we talked about before, but that can lead to excessive penalties for those transactions. So that actually drives up your cost. Whose fault is that? Will they back time to your actions? So are you really taking actions that are effective?

Or in the card industry, people talk about downgrade, like the data you’re sending or not sending or impacting your cost. And sometimes that can be bigger than what you’re paying your acquirer or PSP. Did you send the AVS data or did you send the AVS data a year ago when we launched, but now you have a bug that means you’re stripping out the AVS data on the IOS channel? Did you even know that? No, but that’s actually why the cost is going up. Or you took long to sell. If you’re a retailer or these days you’re not sending a network tokenization transactions, you get an interchange penalty.

And then if you’re using vendors that are helping you with some of these strategies, are the fees that you’re paying them actually effective, or are they just helping you to drive up your costs? As an example, we said before, just brute force retry everything. Is that effective or is it driving up your cost? And that’s just directly linked to the payments, but what is your total cost of payments acceptance? Are you baking in the consequences of refunds or chargebacks or Ill-advised fraud management strategy that are blocking a lot of transactions from people who actually want to buy on your site and are good customers but actually can’t because your settings are wrong? That adds to a much bigger consequence.

And I found a lot of people don’t look at it from a total cost perspective. They’re single down to like, “I’m paying five basis points and I really want to negotiate it to four and now I’m a hero.” That was nothing compared to what the total cost saving could have been if the approach was a little bit more tied to the total cost of payment acceptance. Or even am I being billed correctly from month to month? I think in reality a lot of people in payments are doing things manually on the vendor side and there are errors happening all the time. And are you on top of that or “Hey, it changed. Why did it change? I didn’t do anything different from a volume output perspective.” “Oh, that’s a good question. You’re right. We had an error in the billing cycle or something.”

Yvette Bohanan:

And sometimes these contracts are complicated. There might be volume tiers in them, there might be different things that influence what price point you’re at with a provider, and things get messed up. We see this all the time. We’re going in and doing forensics on this stuff for people and a lot of the forensics are unfortunately manual because they don’t have any recon or any tracking in their system for this at all. So you’re pulling invoices and starting from scratch.

Klas Bäck:

Well going back to what Drew said before, a real scenario. So one team is saying, “We’re going to launch buy now pay later. The way we’re launching that is adding a virtual card.” The virtual card runs on a premium payment method, that was not accounted for for the team that are managing and maybe responsible for the cost, but it drives up your interchange. Now what, everyone is surprised because people weren’t looking at the whole lifecycle and weren’t aware of the same dataset. So it became a cost surprise just because the data wasn’t shared across all the different teams.

Yvette Bohanan:

Yeah. What happened in my debit interchange here, it’s suddenly creeping up month over month? What’s just happened with our customers, who would go back to buy now, pay later right away? That can take a while just to figure out.

Drew Edmond:

I think a big thing too is the variability in the data that you’re getting from. If you’re in a multiprocessor world, there’s a lot of variability in what you get in terms of the granularity of the information on a per transaction basis from your providers. I was talking to somebody yesterday, they were using multiple providers and with one of them he said, “I know that they’re cheaper. I negotiated the contract, I know that they’re cheaper, but when it comes through in the reconciliation, my VP of finance thinks it’s more expensive than our other provider. And so I’m butting heads with them and having conversations, trying to convince them that they’re actually not more expensive.”

If that’s the case, that’s wild. It’s in the numbers, it’s in the reconciliation. What’s going on there? Well, it turns out that they’re incorporating not just the transaction fees when they’re netting out fees and there’s additional services that they’re paying for, but it’s not granular, it’s not available in the reporting. So they don’t know that that’s not transaction level pricing. And so it causes a lot of, I think, problems internally when you have that lack of access to that granular detail.

Yvette Bohanan:

Oh, absolutely. Yeah.

Klas Bäck:

I think in the, going back to the total cost, if you have two vendors and you are trying to compare them to say… what you were saying Drew, “Do you even know that you’re sending them the same data?” Simple question. Very few people can actually answer that quickly or at all. And then if that statement then is true, you don’t know or it isn’t the same data, what are you actually concluding here at all? You can’t, since the consequence of your data you’re sending has maybe a bigger impact on the end cost than what it says in the pricing.

Yvette Bohanan:

Going back to that initial question about how has the role of the payments professional changed in that and the cost question, I think the last time I checked, which has been a little while, Visa, you have interchange and you have assessments, but they have 137 different fees, I think, that could potentially be applied to any transaction or a group of transactions. It is super complicated. And there is calculus now that people are trying to figure out of, “Do I retry and pay that or do I not? Do I tokenize or do I not?” Because there’s a fee if you do, there’s a fee. If you don’t, there’s a fee either way. But what’s the right thing for you to do? How does that affect… My final question. Authorization rates. So somebody walks into the room and says, “Our authorization approvals dropped last week by two basis points, maybe more. Who knows? Does anyone know why?” And everyone’s sitting there in silence.

Klas Bäck:

It then comes back to pretty much all the things that we talked about. Can we establish the data? Is it the right data that we’re looking at? Can we break it down in all of the components that maybe matter? Let’s say if core transaction, can I establish on the totality? But can I establish, actually it’s only happening on this platform or this product or this vendor or it’s actually tied to specific use cases? So it’s only happening in Portugal, or it only happened with one issuer in Portugal, but that happened to be a meaningful part of my total transaction. So it all stems from there. Or it only comes from one of my vendors and only in one country, but not in the other one? Why is that? Not necessarily something I did. Or was it something that I did? How can I establish that?

And if you can’t do that, how are you ever going to answer any of these questions in terms of establishing the benchmark and understanding your consequences and how you can do better. Or even establish, am I doing a good job or not? That we were talking about before. How does this compare to my peers? Even on that level, that is not an easy activity for almost all companies out there. It’s actually very hard. And most decisions in payments end up being not data-driven. Even for extremely large companies, they’re data-driven for everything else. They rarely are when it comes to payments. Strangely enough.

Yvette Bohanan:

Yeah, I think that’s a good way to bottom line everything. And I guess it’s fair to say that if you don’t have a decent handle on the data, if you haven’t understood the fact that this is hard to get to, it’s hard to get to in a usable format, you need to know how to interpret it. What am I looking at? To your example, Drew, if the CFO doesn’t realize how much stuff’s being netted out of the settlement amount to him, you don’t even know how to interpret your settlement amount. And then knowing how things are implemented, Klas, I think you’re… I can’t even count how many times in my career, you too probably, you sat in rooms and people were like, “Oh yeah, we did that code push Thursday, we forgot about that. Maybe that’s the problem.” Or knowing what the word idempotent means, for example.

So it can be hard to get at this within your org because everyone’s got a different slice, a different perspective, a different level of understanding and different data that they’re looking at to form their opinions and see if there’s a problem or not.

Klas Bäck:

Even going back to even more fundamental to your question, did anyone actually monito or how many companies didn’t even realize they had a drop in approval rate? Yes, there’s some people that did discover it and then all the complexity that we just talked about, but most people probably wouldn’t have caught it or if they caught it it is much later. And we’re not talking about two basis points here. It could be 2% like, “Hey, it used to be 87%, now it’s 85. And I don’t know why.” If I caught it. And did I caught it quickly? When I start looking, why is our cost so up? Oh, look at this. Three months ago something happened. I didn’t caught it then, but now I realize something happening. That is a very real scenario for quite a few merchants that we run into over the years.

Yvette Bohanan:

We run into it too, a lot. Yeah. A lot of times that’s the phone call we get, “Hey, we’re trying to figure out why we went from 87 to 85” in your example, and we’re like, “Let’s start at the beginning.” But then when you get them there, when you actually start to put some structure around this and people start to put a little discipline around it and communication starts to get there, you can start to answer some of the questions. And I guess that’s the hopeful side of things, but if you don’t have a decent handle on it, don’t expect to be able to answer the questions.

Klas Bäck:

Meanwhile, as we established it is a very material part of the total performance of your company and the total cost of your company.

Yvette Bohanan:

Yeah, absolutely. We’re asking these questions, we’re talking about this, there’s KPIs and there’s insights into what’s going on. Maybe it’s worth considering the difference between having some basic KPIs, knowing your decline rates or whatever, and getting to a place where you’re actually optimizing. You’re not in that reactive and on your back foot mode, but you’re able to be proactively looking at things and making changes going forward. What have you observed in that evolutionary path of getting from KPIs to an optimized state?

Drew Edmond:

Yeah, I think you’re right in that establishing, at least understanding how to set the KPIs in the first place. And part of that we’ve been talking about benchmarking and understanding what is a reasonable number here to look at in the first place? How am I supposed to set alerts and thresholds and things like that if I don’t know what good and bad looks like? So you need some education there, whether that’s going out and having contacts in your industry or working with folks that can provide you with that. Sometimes your acquirers can or other providers, to help you at least baseline that.

And then as a former boss of mine used to put it “Find quarters in the couch cushion. Let’s start finding ways to improve things tick by tick.” And sometimes it’s larger ticks than others. Sometimes those quarters are real dollars. And I think it’s understanding and learning enough about each individual piece. Are you managing your costs? Are you minimizing risk? Are you trying to maximize your revenue? Are you ensuring your own data integrity of the reports that you’re getting back from your providers? Are you monitoring your own platform performance? So there’s these core categories that you need to start thinking about optimal and where are we at and what can we do? What are the tactics and strategies and tools and things that we need to do to improve across all of those?

Klas Bäck:

Yeah, I think that’s exactly right. I think almost all companies should track more things. They need to do it more granularly in terms of finding where it comes from. They need to find and discover the issues quickly, and they need to establish that benchmark, and they need to be able to react to it and use it for “I can prove I’m making improvements here.” And they need to operate on that data in a way that everyone shares the same data. So they’re making the right decision based on that that actually benefits the company a little bit, to some of the examples that we talked about before.

And there is standard stuff, the major things that many are looking at. And then there are things that are specific to certain industries like retry logic. For subscription merchant or if I’m a marketplace, maybe people can create fake seller accounts and run stolen cards through it and get away with it. That’s a function of the product. So you actually have money laundering and losses and you need to have KPIs tied to payments that are related to that then. But even if you simplify all that down, it also… Any of that data changes over time all the time. So I think if you take any data point and actually plot it on the chart, you’d be surprised what you find. How come my data in the last three days, three times changed in one country for a period of time with one vendor? Where does that even come from? And you wouldn’t see that if you don’t plot the data in more time units. And that on itself is a pretty big contributor to how to drive better performance. Just look at it over time.

Yvette Bohanan:

If you can get to your own data, just your own data and understand what’s going on and what looks normal. I think that’s what you’re really bringing up here, and that’s fascinating to me, what’s normal for us as a business in our business versus what’s out of bounds or out of band. And then having, as you’re saying, Drew, the monitoring and alerts on it and stuff.

I think one of the stories that I would bring up is… Now, I was going to ask you both this too, has there been a surprise moment where you’re sitting in the room and because someone had their data figured out, they were able to do something that was sort of a eureka moment or a game-changing moment? I personally recall we were in a quarterly business review with one of our payment service providers and we were looking at their data and we brought up our own data. They had all these slides and everything about our performance and our auth rates and our costs and everything else.

And for years we would nod our heads in the room and say, yeah, that looks great. And we had been putting a lot of work into our data and we had been talking about it cross-functionally for every week. So this QBR comes up, they throw up these slides on the wall and we look at each other and we pull up our graphs and we say, “We don’t agree with you.” And I remember it so distinctly out of all the QBRs I’ve been in, because it was the first time we were able to say we had a handle on our data confidently enough that we could actually say in the moment, “You need to go back and re-look at that because we don’t agree with what you’re saying here.” We’re being polite, but we knew the data was wrong and it led down a path where we optimized a whole bunch of stuff.

Klas Bäck:

I don’t know if anything is really a surprise because the reason why we started Pagos was exactly for this reason. Almost no one is data-driven around payments. And even the ones who are by the way and say, “Hey, we consider ourselves aware.” I think you find there’s a lot of pennies in between the pillows and in their sofa as well. So even those have more opportunities than maybe they think. They specialize in something very specific and it’s like, “Hey, there are a lot of other good ideas you can borrow from,” or even establish benchmark data to help you find more things and help you find things faster.

Because things change over time. In your environment, in your vendor’s environment, in the broader ecosystem. So it’s a constant ongoing thing. You got to monitor and observe in order to really perfect it. But yeah, people find alot of things all the time if they have the data and they have the help to actually understand what they should do about it. But I also think the variation of what you were saying Yvette is like, “Okay, you’re telling me this data, dear vendor, but what does it mean and what should I do about it? Is it good data or bad data? If I compare it to all my peers. Here they are. Are they doing better or worse?” And often people can’t establish that level, so then it’s also hard to take action of “No, in fact, you’re 10% worse than everyone in your industry category, so you’re leaving a lot of things that you can do so much better that have big impact on your business.”

Yvette Bohanan:

Benchmarking is crazy invaluable.

Klas Bäck:

How do you benchmark if you don’t know what your own data is and therefore what conclusions do you draw from it.

Yvette Bohanan:

Yeah. So getting from KPI to insight, that’s a great example.

Drew Edmond:

And then insight to outcome. I think that’s what we’re talking about too. Taking it all the way to actually making changes that affect your business. My example would be being able to go in and show once you actually start applying some of these filters to the data, you say, “Oh, all right, we’re splitting it between card brands, we’re splitting it between card types. Now we’re getting down to issuer level details. Oh, we found some banks that are sending strange decline reason codes and disproportionately declining transactions at a high rate.” That’s fixable stuff. You can go to your acquirer, they can talk to the bank. All of a sudden, oh, the banks themselves are doing something weird on their end, now we can fix that and the authorization rate goes higher. So that’s like tangible money, tangible revenue, a great outcome when it comes to applying the KPIs because we knew the benchmark, we knew they were below it because we knew the insights of, “Oh, we need to apply these filters to find it. Now we can have the outcome that we’re looking for of improving our metrics.”

Yvette Bohanan:

Exactly. Yeah.

Klas Bäck:

And on that related note, I think is shocking how often uncertainty in the data means people can’t even build the business case that’s strong enough to actually do something about it. Million of scenarios, “Should I do level 3 data? Should I do debit card routing? It’s like, I’m not sure what the impact truly will be. Should I use network tokenization? Is my account updater strategy effective or is it too expensive?” The million questions that people have that is fairly straightforward to answer except they don’t have that data, therefore they can’t answer it, therefore they can’t build the business case. Therefore, to your point, Drew, they actually can’t take action on it.

Drew Edmond:

I think, Klas, you brought up something really important too, and it doesn’t end. It’s not a moment in time where you can say, “I’ve done it. I’ve optimized payments. My work here is done.” I wish it was that easy. You can’t hire a temporary team and it comes in and you fix it. No, things change. New technology comes. Now we have network tokens, now we have 3DS, now we have eight-digit BINs.

Klas Bäck:

Network bulletins every quarter.

Drew Edmond:

Exactly. So you got to stay on top of it, and you have to have a systematic way to grow in your education of it and the frameworks that you’re building to adapt to the changes in the ecosystem.

Yvette Bohanan:

Yeah, and it’s a great point. Great point.

All right, Klas, Drew, it’s that special time when we have to wrap things up. So thank you so much for joining me on this episode. I really appreciate it.

Klas Bäck:

Well, thanks for having me. It’s always fun to talk payments.

Drew Edmond:

Thanks Yvette.

Yvette Bohanan:

Thank you both. And to all of you listening, maybe this episode has inspired you to consider your payment data differently. Maybe you’ll share it with some of your colleagues and start a cross-functional meeting and start digging into your own data. Maybe you’ll have this to give to someone so they have a little more insight into why these questions are difficult for you to answer and what you might be able to do to start down the journey of being able to answer them. We hope so. As always, thanks for joining us and until next time, keep up the good work. Bye for now.

If you enjoy Payments on Fire, someone else might too, so please feel free to share this podcast on your favorite social media outlet. Payments on Fire is a production of Glenbrook Partners. Glenbrook is a leading global consulting and education firm to the payments industry. Learn more and connect with us by visiting our website at glenbrook.com. All opinions expressed on our podcast are those of our hosts and guests. While companies featured or mentioned on our show may be clients of Glenbrook. Glenbrook receives no compensation for podcasts. No mention of any company or specific offering should be construed as an endorsement of that company’s products or services.