JOHN COLLISON: Good morning and welcome to day two of Stripe Sessions. And in my defense, cookies are an important part of the future of payments, especially with the demise of third-party cookies. You know, they’ve been critical for personalization, which we’ll talk about in a bit. And it’s not just cookies. Nearly all of the technology trends you encounter here in Silicon Valley have some important payments impact.

Maybe the biggest example of that is AI. AI agents are likely to change commerce in a big way. OpenAI recently showed this demo of ChatGPT doing your groceries on Instacart. It’ll suggest meal plans and pull recipes and drop everything into your cart that you need. And you know, everything, sales, marketing, product… they’re going to be very different in a world where your end customers aren’t directly browsing your website but they’re talking to an AI agent that’s doing tasks for them. It’s going to introduce a very new and different set of challenges. So we think the future of technology is in many ways the future of payments, but there [are] also a whole bunch of payment-specific trends that you all should be across. And that’s what we’re going to focus most of our time on here today.

You might be familiar with this quote from William Gibson: “The future is already here, just not evenly distributed.” And this is actually very true in payments. So for the next 45 minutes, we’re going to take a look at the pockets of payments innovation all around the world and what they might mean for you. We’ll look at why crypto is back, why you’re maybe underestimating BNPLs, the future of the checkout page, and a lot more.

So our first pocket of the future is real-time payments, or RTPs. The domestic digital payment systems that are low cost, secure, and settle instantly—the kind of payment systems you bring home to your parents. You’ve probably heard a lot about these. RTPs are built of two tightly integrated components. You need both. Firstly, you have a rail for bank transfers that settles instantly. And then secondly, you have a payment method that’s built on top of that rail. And the payment method part carries all the consumer UX and scheme rules and all the nifty features you’re used to recurring payments or something like that. It’s basically where the government set out to design Venmo or Cash App and actually did a surprisingly good job of it.

In some of the world’s largest economies, RTPs are becoming the most popular payment methods. Look at UPI in India. Maybe the most famous example. Cash has long been king there, but not anymore. UPI launched in 2016 and already has over 300 million monthly active users and 130 billion annual transactions. And this explosive growth has correlated with cash transactions actually declining by 12 percentage points over the last five years as the government has run a project to try and wean the country off of cash.

Another example is Brazil, where there [are] 200 million active credit cards. But also, we saw PIX launch four years ago, and there [are] already over 150 million PIX users. And Brazilians are more commonly paying with PIX. Last quarter... last year, excuse me, quarterly PIX transactions began to outstrip card transactions.

And this trend of RTPs taking off… it’s not just confined to emerging markets. Let’s look at an example in Sweden. So the most popular way for Swedes to buy their fermented herring is with Swish. Swish usage is now double that of credit cards. And to give you a concrete example of how these products work: if we just look at the checkout flow for Swish… so this is starting on the Stripe page, we can select Swish we’re going to be bounced over into Swish here. We just tap “paid”... this is our bank app…we enter our security code to authenticate. And that’s it. That is the transaction done. It’s not what you’d normally think of as a bank payment method. You might think they’d be kind of clunky. It’s quite slick.

So along with bicycle lanes and banning emails after 5:00 p.m., RTPs are all the rage across Europe. We’ve seen schemes from Bizum in Spain to iDEAL in the Netherlands. The rollout of a pan-European instant bank settlement layer, SEPA Instant, is surely going to accelerate the number of RTPs in Europe. So if you are selling your products around the world, the emergence of all these RTPs means the payment method preferences of your customers are changing. You need to be future-proofed for this new world of payment method proliferation.

What about here in the US? So you’ve probably heard a lot about FedNow, which is live with over 600 participating banks moving money on these new rails. It’s awesome. And you hear FedNow compared to real-time payment schemes like the ones we just saw. But that’s actually a bit of a category error because FedNow is a rail for the instant settlement of bank transfers, but there’s no payment method on top. So it’s probably better to compare it to Zengin in Japan or Faster Payments in the UK. These have been around for quite a long time, but they are just vanilla bank transfer schemes. And to be clear: they’re still useful. There [were] a few billion transactions settled on each of those systems last year, but it pales in comparison to something like PIX, which processed 43 billion transactions in 2023—an order of magnitude more. So you can probably expect FedNow to be similar to those. It’ll be useful for your push payments, paying salaries, suppliers, but we don’t think it’s going to change consumers’ payment preferences in the US anytime soon.

So that’s RTPs. They add to the complexity of running a global checkout. It’s more payment methods for you to keep on top of, but of course, if you’re on Stripe, adding more payment methods to your checkout is easy and you’ll definitely appreciate the faster settlement. But the thing about RTPs is they’re part of a broader trend shaping the future of payments which we see… which is that payments are becoming more political.

So this is the ARA Libertad, a ship in the Argentine Navy… I have a lot of questions actually… But a decade ago it was seized… not by pirates, not by the British, but by Elliott Capital Management, a New York hedge fund that held bonds on which the Argentinian government had defaulted. The US Supreme Court ruled that Argentina’s creditors could issue subpoenas to banks to trace the county’s assets all around the world. And you know, a Navy ship going missing, it tends to get your attention. It’s the kind of thing that gets governments wondering, “Where exactly do our payments flow through again and what are the rules?” The payment systems we all use every day are an essential utility for modern life. As a result, they offer powerful leverage to advance or defend political agendas. In geopolitics, we’ve seen governments go to new lengths to use payment systems to advance foreign policy objectives.

For example, the exclusion of certain North Korean and Russian entities from the SWIFT network. As a result, governments around the world are trying to gain sovereignty over their payment systems. Domestically controlled payment systems, like the RTPs we just saw, are a step by governments to run in-house their most heavily used day-to-day financial infrastructure. And RTPs aren’t the only example of countries trialing new payment systems. Many countries are already conducting live experiments with CBDCs. For a good number of these experiments, a key objective is to expand the use of their currency versus—for example, the US dollar. Additionally, many CBDC experiments are looking to create a new unified standard for cross-border payments. We’re also seeing the growth of alternatives to the SWIFT messaging system, such as Russia’s SPFS and China’s CIPS, which are explicitly targeted at sidestepping the dollarized economy.

Even within domestic politics, we’re seeing payment systems being drawn into the political fray. Within payments, we’re seeing new pressures on financial service providers to either remove or reinstate entities caught up in controversies. Advocates will say this is an effective way to limit harms that kind of outrun the justice system. But opponents say this is cancel culture hitting the financial system with the latest enemy of social media being on the chopping block. Nigel Farage had his accounts at NatWest closed apparently because they were worried about associating with him. When Canadian truckers took over Ottawa to protest COVID restrictions, the Canadian government pressured banks and payment companies to cut off donations. But then US officials took the opposite view and they announced they’d investigate payment companies that refused to process payments for the truckers. In Operation Choke Point… that was the name of it… Operation Choke Point… the US government pressured banks to not provide services to firearms dealers and payday lenders and businesses like that. Then a different administration tried to prevent banks from restricting services to those very same businesses. Before digital payments, when everything was cash it simply wasn’t possible for payments to be a choke point. Whereas now, in a world of digital payment providers and social media controversies payments itself can quickly become a divisive issue.

Politics and payments is one of the hot topics you tend to hear about if payments is in the news… and I mean, payments in general is not newsworthy. But our next topic is one that you will occasionally see headlines about, and that’s payment costs.

Following payment costs is a bit like planning a wedding. There’s a lot to orchestrate and somehow, the prices seem to always be going up. So this is a chart of payment costs for businesses in the US and as you can see, there’s kind of an upward trend over the past five years. But you probably saw this big news just a few weeks ago… a settlement between merchants and the card networks over interchange fees. This case has been running since 2006, so it’s taken a while. But the two major card networks have struck a deal intended to reduce payment costs in the US. What’s being proposed in this settlement—it’s not approved yet—but what’s being proposed in this settlement mimics two of the most common approaches we’ve already seen around the world to lower payment costs. So first is the obvious one, just a straight up cap. That’s what tends to get all the headlines. But the second is enabling price competition for payment methods in the form of surcharging.

So let’s take a look at how both approaches are faring. First, regulated fee reductions: do they actually sustainably cut costs? You might think the answer is obviously yes. But when we look at markets that have regulated lower fees, it’s actually a bit of a mixed bag at times. So take the UK, which introduced domestic interchange caps in 2018. You saw payment costs fall, but then they started to climb again when some of the Brexit-related changes kicked in. So that didn’t really last as intended. New Zealand, on the other hand, introduced exchange caps in 2022 and those reduced costs have sustained. Now you might say that’s only two years of data. For another example, we can look at regulated debit card processing in the US. The Durbin Amendment regulated debit card prices in 2012, and that has held debit card prices down for 12 years. So regulated cost reductions can sustain, but it’s not an ironclad guarantee. If the card network deal is approved by the courts, we can expect to see fees come down by a handful of BIPs over the next five years.

But then, like I said, the second approach to lowering payment costs is surcharging. This means you’ll have the power to charge different prices to customers using different cards. Rewards cards might feel a bit less rewarding if customers are suddenly paying more for them. We don’t have to hypothesize about how this will work. We can just go ask the Australians who’ve had surcharging rules in place for years.

So what can we learn from the Australians? Well you can learn that government spending can get out of control. This is actually a spending scandal in Australia. But sorry, I meant, what can we learn on surcharging specifically from the Australians. We learned that 45% of consumers opt to use lower-cost cards rather than incur extra surcharges at checkout—so much so that the median surcharge amount has actually gone down from 2% in 2019 to 1.5% now. That’s just surcharging is really working in shifting behavior. You can already see this behavior happening at a smaller scale in the US where businesses look to offer consumers lower-cost payment methods, even outside of cards.

Take Airbnb. They’re using our Link bank account product, our bank payment product that Will showed you yesterday, where they’re offering discounts to customers making longer-term bookings if they use it. So you see here, you save $192 if you pay with your bank account. Yesterday, you all got your Uber discount code if you paid with the bank account on Uber. Similarly, Uber is in the process of incentivizing their customers to use lower-cost payment methods with Link.

So if payment costs are a major consideration for your business, you might need to start thinking about surcharging and offering customers different payment methods. And experimentation and testing will be key as you trade off amongst all the cost and conversion and authorization factors. But we’re excited about this because it puts you in the driver’s seat of managing your costs. So that’s payment costs—sometimes going down, usually going up, maybe about [to] start going down again… especially if you want them to.

It’s now time to hear from our first guest of the morning, and to talk to us about the future payments I wanted to bring someone who has been a direct catalyst in bringing the future about. Alex Rampell is the cofounder of two payment companies, TrialPay and Affirm. Today, he’s a general partner at Andreessen Horowitz, where he sits on more than a dozen boards in investing companies across fintech and AI and so much more. So welcome, Alex.

JOHN COLLISON: So I introduced this with the framing that... The disruption is here, just not evenly distributed. And if you’re looking around the world you can track that quite well. Are there disruptions that you are looking at that you think will be kind of significantly more impactful in 5 or 10-year time?

ALEX RAMPELL: Well, I mean, I know you mentioned AI, and AI is the hot thing right now. And I’m sorry for being a VC that’s talking about AI, or talking about wine, or talking about cycling,or whatever VC’s are talking about. But here’s how I think about why AI is actually very impactful and how it relates back to payments. So, vertical SaaS became a really big thing and part of the reason why every restaurant 40 years ago should have used software. But, in order for them to use software, they would’ve had to buy an IBM AS/400 or something, install on-premises software, [and] have legions of IT people. That was just not really possible and it was all too expensive. Cloud made something like Toast possible. But actually what made Toast work as a business was fintech, and specifically payments, because now I can say you can use software as long as you have a web browser. Everybody has a web browser, great, now you can use my software. What I’d like to do next is make money, and I can only charge the restaurant $100 a month. There are only so many restaurants; that business model doesn’t work. Oh wait, I can actually embed payments, I can embed payroll, I can embed these other things.

So what happened was a lot of vertical SaaS models that didn’t seem big enough got big enough when you embedded payments. And the same thing is, actually, I think about to happen with AI, where a lot of what happened in software was… take a company like Workday that we were talking about earlier. Workday basically digitized the HR filing cabinet. It didn’t help you hire more HR people. So a business that, I mean, Workday was a big business. It’s a $50 billion company, but there are a lot of companies where it’s like they just don’t seem big enough to turn into a business, so nobody makes them. But if you can actually monetize, step one was sell it in the cloud so that anybody can use it. Step two was embed financial services. That was the giant leap. I mean payments embedded with financial... sorry, payments embedded with software became big. But step three is, I don’t know. Take a business where they run a Microsoft Excel. There’s probably not that much software to write, but it’s really hard to hire the 7,000 people that use Microsoft Excel all the time. That is now a business model. And I think a lot of that will be combined with financial services and integrated financial services.

JOHN COLLISON: Okay, so you think AI will be kind of the next wave of that if we look at. I agree with you, by the way—and we’re going to talk about vertical SaaS much more in a little bit—that fintech made vertical SaaS a better product because you have this new revenue line, presumably AI won’t be a new revenue line. Are you saying that it will make vertical SaaS more possible, more compelling, but you’ll still charge for the software, or will the revenue model be different?

ALEX RAMPELL: No, I think you’ll probably still charge for the software. What Tom—Tom is the CFO of Mindbody—we were talking about a business called PoolTrac, which is like the most obscure vertical SaaS company in the world, but it’s for companies that maintain swimming pools, and how many companies are there on Earth that maintain swimming pools? There aren’t that many; that market can’t be big enough. Then they won’t pay $10,000 a month for software. They might pay $50 a month for software to do CRM, but once you embed payments...

JOHN COLLISON: There’s no size for software today and AI brings that down.

ALEX RAMPELL: Exactly, exactly.

JOHN COLLISON: I like that. And speaking of revenue models, you have this view that way more companies should do their own version of, kind of, Target’s Red Card. Can you explain that idea to people?

ALEX RAMPELL: Yeah, so if you... Target’s a public company, so if you look up their 10-Q or 10-K, one of the KPIs that they point out is what percentage of their payments were processed effectively on their own network. Because Target, I think their second or third cost is payment processing. Target has something like a 2% net margin. So you take their net income, divide by the revenue, 2%. If they didn’t have to pay 100 to 200 basis points of payment fees, like guess what? They would make almost twice as much money. Highly impactful. So what they did was they said, “We will give you lower payment processing costs,” actually, sorry, for the consumer they are giving a 5% cash-back rate if you sign up for the Target Red Card. The Target Red Card is two things. It’s a co-brand card. I mean Macy’s today is not doing great as a business except for one line item, which is their co-brand card. United Airlines makes lots of money on their co-brand card. Co-brand cards have been known as a very profitable thing for a long time. So they have one of those, which is issued by Toronto-Dominion Bank.

But the real Red Card thing that they do that’s interesting is [that] it’s kind of like this abstraction layer against your bank account. It’s exactly what you were talking about with Link to draw directly from a bank. And they basically built, for lack of a better term, an abstraction layer because you go pay with your Target Red Card—it’s a debit card but it’s a closed loop debit card—and then it just withdraws money from your bank account. They had to build a lot of complexity there, because how you map chargebacks in the credit and debit card world, which are pretty well-defined, into “there [are] nonsufficient funds in the bank account” or somebody [asking], “I’m alleging theft on my Target Red Card; how do I deal with that?” Because Bank of America doesn’t want to. I want my money back in my Bank of America account.

So what I found compelling about the Red Card is Target wants to lower their payment processing costs, which is obvious, everybody does, but the way that they did it and the fact that they got to 20% share of all of their consumers using this payment instrument was extremely interesting. And the way that I’ve thought about this is for repeat billers, like not everybody can be Target because Target, you might go to twice a week.

JOHN COLLISON: Yeah, the frequency, yeah.

ALEX RAMPELL: And what you’ve started seeing with a lot of the frequent billers, and Uber is like that too. Uber is not like I need to buy it every day. It’s not like my electricity bill, but my electricity bill, my Comcast bill, like all of these things that I pay for on a regular basis they’re all trying to incent people. They’ll say, “I will give you $5 right now if you switch from credit card to bank account.” And that’s starting to work. So the ones that have high frequency of interaction with the same customer, I think Red Card is a really good... it’s a harbinger of what might come.

JOHN COLLISON: So in the closed-loop world, everyone likes to talk about Starbucks. Do you think maybe Target is a better example for people to look at?

ALEX RAMPELL: Yeah, yeah, I do.

JOHN COLLISON: And last question. We’re thinking a lot about, kind of, permissioning and identity in these topics and obviously that changes a lot in the world of AI. I know you’ve been thinking about that topic as well. How should people here be thinking about, kind of, permissioning and identity in this world?

ALEX RAMPELL: Well, I put out a post a long time ago. It’s kind of hard being right but too early which is the same thing as being wrong, by the way. It’s a nice way of saying I was wrong. But when the Apple Wallet came out if you think about what you get to permission on your phone right now, you install an app and it says “Hey, I am the app, do I have access to your contacts?” “Yes or no?” And you, John, say like, “Yes, you do,” or “No, you don’t.” Location, yes or no. Photos, yes or no.

You already have a handful of cases right now where your ID card is like... I think there are four or five states in the country now, or in the US, where you could provision your DMV-issued driver’s license into your Apple Wallet. I assume that Google has something very similar. You can imagine a world in the future where I want to apply for a new payment instrument. Now in the future, that might be a button, like there might be an App Store for the Wallet where the way that I apply for anything is like... but the thing on permissioning is... like the Social Security number that I have is in my Wallet, my driver’s license is in my virtual wallet. Those will have the same kind of permissioning or permissionings as you see right now for... location, photos, and the like.

JOHN COLLISON: But you also made the point that maybe just clicking “yes” in a dialog where it is the user is not enough to actually treat as [a] kind of permission these days. You know, your example was that... Your casino in Vegas example.

ALEX RAMPELL: Yeah, well, I think this is the strange thing, is things move from the offline world to the online world. So there are really three things that you need to know from an identity perspective. Number one: is this identity real? And it’s actually quite easy to create a fake identity. We had a lot of problems with this at Affirm. I’m sure you do as well. So, is this identity real? Then number two is: is my interlocutor, the person that I’m interacting with—is that the owner of this identity? And that’s also very hard, and we actually have pretty good technology...

JOHN COLLISON: Ask a bouncer in a college town where they’re just trying to distinguish between members of the family.

ALEX RAMPELL: Exactly, exactly. So, and then part three which nobody has really thought about yet is, is the person of sound mind to have made this decision? So if I go on Robinhood and I buy a bunch of YOLO calls that expire today with my entire net worth, and I’m sloshed, right? I’m completely wasted. If I went into an old Charles Schwab office in 1970, they wouldn’t let me do that, right? And it’s not like Robinhood wants to let you do that either, but you don’t have a way of actually figuring out, like, is the person of sound mind to make this decision? In Vegas, they want you to lose all of your money, but they also don’t want to go to jail. And there are rules about this they will not serve you your 15th drink and let you YOLO all of your money on double zero in roulette. So I think that’s another... As everything that happens in a financial transaction and the offline world starts moving online, how do you, not to be too skeuomorphic, but how do you encapsulate and decision some of these things?

JOHN COLLISON: I think that’s a very astute comparison, which I think is, people talk about identity online, they generally mean the first one of “do you have kind of valid ID?” And it’s not, is the person actually your kids using your phones, or whatever is a big example of the second one, and then is the person of sound mind. Anyway, we can go for hours more, but we only have a brief amount of time. Alex, thank you for coming today.

ALEX RAMPELL: Of course, thank you very much. Yeah, thank you, everyone.

JOHN COLLISON: Next up we are going to talk about crypto. Now if you’re a nightclub owner in Miami, you might have noticed that business has been a bit slower than usual. It’s been a rough few years in the crypto market, and crypto’s the only part of the financial world where you can get 12-year-olds making themselves millions by drawing a whale wearing a top hat and people making and losing billions on Dogecoins. But if you’re just skimming the funny headlines and chuckling and moving on, you’re missing out on all the genuinely exciting progress that’s been happening in crypto, and you owe it to yourself to dig in properly.

Our history with crypto goes back a long time. You know, 10 years ago, we became the first major payments company to support Bitcoin as a payment method. At the time, Bitcoin was a pretty terrible payment method. Transactions took ages, they’re very expensive. So in 2018, we actually turned it off. But amidst all the trials and tribulations and Shiba Inus and crypto kitties, there [have] been a lot of technical improvements happening in crypto. I’m talking about three things specifically.

Firstly, transaction settlement times are no longer comparable to Christopher Nolan films for length. Many blockchains are now measuring transaction times in the milliseconds. Secondly, transaction settlement costs are no longer comparable to Christopher Nolan films for budget; it’s now perfectly affordable to transact on chain. Finally, and this is a big one. Stablecoins are becoming, well, stable. They’re regulated, they’re backed by real assets, they’re being used a lot. Look at this chart of the total being transacted by stablecoins. If we are in a crypto winter post-2021, nobody told the stablecoins.

So what does this all mean? Well, it means crypto is finding real utility. We’re seeing folks use it to store value most notably when their domestic currency crashes. You don’t see this in the US; it’s a very good currency here, but that’s not the case everywhere. Look at Turkey; you can see that the price of the Turkish lira has fallen over the past few years. The lira’s trading volumes against stablecoins have consistently increased. With transaction speeds increasing and costs coming down, we’re seeing crypto finally making sense as a means of exchange.

So, as something for you, we’re excited to announce that we’re bringing back crypto as a way to accept payments, but this time with a much better experience as stablecoins. Do you want to see how it works? Alright.

So this is a little preview. Patrick’s not the only one who can leak things. This is the Stripe Checkout you know and love, and you know it’s got all the cool functionality like... again, a big one for us is localization, where it looks different. You know, you see here depending on where the customer is coming from it’ll give them that really nice high-converting localized experience. But what’s this new functionality over here on the right, we can pay with crypto. So we’ll tap “pay” here, we’ll go to a confirmation screen, and we’re on Hyperion the ghost of Christmas past. So this is our confirmation page. We passed in the email address for the checkout. Here we’re selecting what chain we want to pay with. We’re going to pay with Solana. Here we’re selecting what wallet we want to use. We could use Coinbase Wallet or MetaMask. I’m going to use Phantom for the purposes of this demo.

So when I pay here, it’s going to bring up some UI from Phantom, which is the crypto wallet that we’re using. So this is the Phantom UI here. This is not Stripe UI. You see, here, we’re confirming we’re going to pay with Solana. The network fee, very cheap. As we discussed, network fee’s coming down $99. Now again, you’re probably used to a crypto transaction. You’re expecting me to bring out another guest here and we’ll do an eight-minute conversation while we wait for the transaction to clear. But that is old crypto. You ready for the new crypto world? You have to look very closely and not blink, you ready? That’s it. We’re done. That confirmation page was not like, “This has been entered in a queue. We’re going to now go do something.” It has now been posted to the blockchain already. So if I go into my Phantom Wallet here and go into the Explorer, you see here sent at 9:58 today, we can view on the Solana Explorer. This is I think the Explorer just loading… it is on the blockchain. Here we go. So you can go to this URL yourself if you are eagle-eyed and see this is a real posted transaction to the Solana chain. Similarly, we can view it in the Stripe Dashboard. So if we click in here to our 9:58 a.m. payment, you see here we’ve made a payment with crypto and again, Stripe handles everything for you. You just got $98 into your bank account. But that is crypto coming back to Stripe. That will be coming later this summer. Let us know if you are interested in piloting this. We are super excited about it.

Okay, we are going to move on to one of the most debated topics in payments innovation and that is buy now, pay later methods, or BNPLs. Do they drive up conversions and average order values or are they just a more expensive way to make the same sales that you would’ve made with card payments? We’re actually in a reasonable position to test this because we have no horse in this race. But we get a good view of the data so we can put it to the test. Here’s what we found. We did some trials where Afterpay was displayed to consumers, and we had a holdback, and the businesses using Afterpay globally saw a 2.3% increase in revenue. Certain businesses could see a much larger increase, depending on the customer base. So that’s settled: BNPLs are definitively good for business.

And you might think, these are existing sales that would’ve happened otherwise, they’re just switching over to a different payment method. They’re not. We dug into this shift in share of wallets and the test showed that BNPLs are flipping uncompleted transactions into purchases. In our most conservative estimates, 61% of that lift is coming from net-new sales. So you might be wondering, ”Really? What explains this behavior?” But as with many strange phenomena, it’s the pesky kids, the Gen Zs… they just get the ick when it comes to credit cards. The vibes are not there. Credit card usage amongst Gen Z is at 42%, way lower than the 71% amongst Gen X, and they’re using BNPLs in every aspect. They’re spending [on] shopping, utility bills, maintenance. All as an easy way to track spending without traditional credit cards.

So if you want to sell to the folks on the right side of 35, you are going to need BNPLs. You might wonder, what should your BNPL strategy be? Well, there’s a universe of BNPLs out there each with its own kind of specialization in terms of geography and payment type, and your customer’s preferences for specific BNPLs will vary. So you want the ability to support the provider or providers that will best perform for your customers. You want the ability to do A/B testing on your checkout like we showed you yesterday. So that’s BNPL. One of the biggest changes we have seen in B2C commerce. A very important part of selling to your next generation of customers, and statistically, significantly valuable.

On the B2B side, the trend that we’re most focused on is what Alex just talked about: the rise of vertical payments and vertical SaaS companies doing payments. So this is a big change. You know, a few years ago, your local gym got its payments from a local sales rep of a merchant-acquiring bank who’s out there, feet on the street going door to door. They had their reader in hand and they went to the gym and then they went down the street to the barber and then they went down the street to the church and they were kind of selling generic payment services to each one of those businesses. Everyone was going to the bank for payment services.

Then that all changed. That same gym today will get their payments from Mindbody. The barber would get their payments from Squire. The church would get donations from Tithe.ly. I guess the pool cleaning company we get on the slides will get their pool cleaning software from PoolTrac. And these are all SaaS platforms that provide a ton of functionality around CRM and scheduling and bookings and all the likes of that and other bundling payments and other financial services. And Mindbody is in a way better position to offer payments and financial products to a gym than, you know, Bank of America is because they understand this particular business, its dynamics. They can build the payment products that this kind of business actually needs. And we think this is one of the most important and underappreciated changes in the world of payments. Payment systems used to be generic and horizontal, and now they’re being tightly tailored to each business type. So let’s speak to someone who helps drive this change. Please welcome Tom Aveston, the CFO of Mindbody.

JOHN COLLISON: Tom, I understand you’ve been the CFO of not one but two fitness companies. So are you yourself a fitness nut?

TOM AVESTON: Well, the issue is I have two little cost centers—I mean children—at home. So while I used to be a bit of a fitness nut, I’m afraid all I get to work on these days are my dad arms.

JOHN COLLISON: I see. So an exercise in prioritization. Can you tell us a bit about... We talked about, well Alex talked about vertical SaaS and the journey into financial services, we talked about it a bit here. Can you tell us about Mindbody’s journey in financial services that might help ground people a bit?

TOM AVESTON: Yeah, certainly. So, Mindbody started now over 20 years ago providing software to the fitness and wellness industry, and we added payments.

JOHN COLLISON: You guys are the big kahuna in the fitness and wellness industry.

TOM AVESTON: We have plenty of competition, but yeah, we’ve been at it a long time.

JOHN COLLISON: The lawyers of Mindbody LLC.

TOM AVESTON: Yeah, so we added payments a number of years ago, and that allowed us to build a foundational layer that we’re adding more and more financial services on top of. But it all started with payments, and I think payments is the enabler of the business in the box or the operating system that Alex touched on earlier.

JOHN COLLISON: Do you see the arc of vertical software companies such as yourself following this thing we described which is many starting out providing a software layer it kind of becomes... It would be crazy not to provide payments because you have a much better ability to have a much better offering there and your customers are asking for it and then you start getting into things like lending and stuff like that. Does that arc match with your experience or has [it] been different in any way?

TOM AVESTON: Yeah, I think it’s exactly as you said earlier, right? Vertical SaaS is in a much better position to provide payments to provide financial services to their customers because they know their customer’s businesses much better than a Bank of America or a JP Morgan. But also because the various processes and workflows can be deeply integrated. If you’ve got different solutions for different elements of your business then you’ve got to reconcile those at the end of the day and that creates friction, it creates the possibility for translation error, and it adds a ton of cost as well. So by doing it all in one place, a business in a box, an operating system, a one-stop shop, whatever you call it we think it reduces that friction. It reduces cost and ultimately provides more value for our customers.

So we started with payments, we have a capital product—a lending product today—we’ll be adding banking payroll, insurance, potentially collective purchasing. We see a ton of opportunity here and in fact, that’s why we were so excited to hear Will yesterday and all of the new products you announced, right? Because it’ll certainly be a component of our offering in the future.

JOHN COLLISON: Yeah, no, on embedded finance. I mean, we have been investing very heavily first in Connect offering payments for platforms. But again, we find that offering financial services to the end businesses tends to just work tremendously well for those platforms. So that’s why we’ve been investing as heavily as we have.

Your integration point is an interesting one. I guess I didn’t mention that in the setup but you guys already handled the bookings for these businesses and so it’s kind of crazy if you got to the very last step of that booking and to pay for it you were turfed out to something else, you’re an integration problem. So it just feels like the integration is why it makes payments a slam dunk.

TOM AVESTON: Yeah, and you know, some of our customers don’t use our payments. It’s a very small minority and they’ve got a ton of work at the end of the process to reconcile all that payments information with their scheduling reservation information and then back into their bookkeeping and accounting. It’s a ton of work.

JOHN COLLISON: How did Mindbody become the moderately-sized kahuna in a very competitive industry? How did you guys get to the position you’re at now?

TOM AVESTON: A player in a very large market, yep. Yeah so, you know, the software, the integration of that software, the removing of friction and kind of driving efficiency that’s an important component of it. That’s very similar to a lot of other successful vertical SaaS companies.

I think one way in which Mindbody has differentiated itself is also by offering broad consumer distribution. That’s super important to any customer that is looking for new avenues of revenue growth. So we have millions of consumers every month using the Mindbody marketplace app and they’re discovering new Mindbody merchants through that app. Two years ago now, two and a half years ago Mindbody acquired ClassPass and that’s just another vehicle for that consumer differentiation. I think that sets us apart from other vertical SaaS providers who are just focused on the software and the payments and maybe other financial services but don’t provide a way for their customers to grow their business. Indeed now, a large proportion of Mindbody customers make more revenue from the new customers that discover them through the Mindbody marketplace than they pay us in SaaS fees. That we think is an incredibly attractive value proposition.

JOHN COLLISON: So customer acquisition has been a huge part of the differentiation for you guys where every business wants to grow. Do you think that’s a generalizable lesson for vertical software companies in other industries?

TOM AVESTON: Possibly. Consumer distribution is hard, as you know. It’s expensive, so I don’t know if I’d necessarily recommend that every vertical SaaS pursue that. You have to have good software. You have to have good payments. And I’d start with that, and maybe you grow to consumer distribution over time. I mean, prior to Mindbody, I worked at ClassPass and we spent hundreds of millions of dollars building the consumer distribution that we have today. So it’s not an easy task.

JOHN COLLISON: It’s a big investment. You guys have so much opportunity ahead of you. You could expand internationally, you could go deep into financial services, like you described. How do you prioritize?

TOM AVESTON: Yeah, it’s a great question. So I’d say that we are focused... ClassPass operates in 30 countries today. Mindbody has customers in over 100 countries. We are focused on product investments that can scale globally. I’d also say that, as with many late-stage companies, there is a tension between investing to drive long-term revenue growth and short-term or shorter-term EBITDA generation. So if we look at the investments that we’re making in the near term, [they’re] product investments to support our customers in countries where we already have critical mass today on the Mindbody side. On the ClassPass side, we’re interested in untapped markets, and Japan is a great example of that. We’ll be launching in Japan later this year, but also in those smaller countries, where we can service consumers in those countries with resources that we already have in the region. Yeah, that’s how we are managing that tension.

JOHN COLLISON: Makes sense. Tom, thank you so much for being a partner. We really appreciate it and thank you for coming here.

TOM AVESTON: Thank you very much.

JOHN COLLISON: The last topic I want to touch on is something that is a hot debate at Stripe: personalization.

So things used to be not that personalized. Think about network television. Some poor executive had to choose amongst thousands of programs what would work best for the masses. JR coming back to life on Dallas—that was the global optimum apparently. Compare that with our new personalized world. You have the window into your soul that is your TikTok feed. All those endless ASMR videos of the credit card imprinting machines just shoomp, shoomp, swipe, that’s what I got. Now you have a personalized local optimum that is so much better for you and it outperforms that global optimum, and nearly everywhere you look this trend is happening from music curation to advertising personalization outperforming the old global optimums more TikTok For You pages and fewer sitcom reruns.

Our belief is that checkouts fully personalized will drastically outperform the current static versions. And so we’re investing in exactly this. Yesterday we showed our Optimized Checkout Suite, for which we’ve industrialized the process of testing the tiniest tweaks to checkout surfaces so you can learn what improves conversion and by how much. And we showed you how seemingly basic personalizations can deliver giant revenue gains, like [how] presenting price in your customer’s local currency can increase your revenue from international sales by 17%. But there’s so much more that can be done to personalize checkout experiences.

Imagine a dynamic checkout that in real-time analyzes quantities in the baskets and customer-end context and compares it to millions of other checkouts. Your checkouts really should know the difference between idle late-night browsing of athleisure to frantic attempts to secure tickets to the Cowboy Carter Tour post haste. Today, 70% of online checkouts are abandoned. So we think that checkout personalization has a very long way to go with enormous revenue upside for you. That’s why we are continuing to invest heavily in our Optimized Checkout Suite.

So today, we promised you pockets of innovation from around the payments industry that we think are worthy of attention. We’ve sped through a few different topics. We talked about real-time payment methods and how they’re winning in their domestic markets, how you should expect more political complexity in payments, how payment costs have been rising, but you have the tools to control them. How crypto is back on Stripe, at least. If you want to sell to Gen Zs, you’re going to need buy now, pay later. How we talked to Tom on how it’s only the beginning for vertical payments and kind of vertical financial services, more broadly. I think that’s the big trend… it’s not just about payments. Then checkouts in the future, we think will be much more personalized and perform vastly better as a result.

You might notice a theme here that the biggest disruptors in payments are not coming from the US. Regulatory changes, RTPs, BNPLs, the most significant shifts in payments are happening elsewhere first. So what does that mean for all of you? Well, in many industries, it’s still true that you come to San Francisco to see the future, but your lived experience here or anywhere in the US likely is not exposing you to the bleeding-edge future of payments. For that, you’re going to have to cast your net around the world to see what’s happening. You’re going to have to travel figuratively and literally to seek out what’s new and what can help you with your business. But the good news is that Stripe itself is 8,000 people all over the world obsessing over every little detail in the payments landscape so that you don’t have to.