Stripe’s Japan market entry did not go according to plan.
Things worked out worked out well in the end, but they did not go according to plan. Stripe is one of the world’s largest payment processing companies, but they remained flexible and agile enough to take advantage of some of the surprises they faced in Japan.
Today we sit down with Daniel Heffernan, the Japan head of Stripe, and he walks us through what happens when a technically sophisticated and streamlined FinTech company comes face-to-face with the very low-tech and slow-moving processes that make up FinTech in Japan, and how they made it all work.
They faced complex, lengthy technical specifications delivered in three-ring binders and un-copyable, printed documents, and they dealt with the Japanese aversion to integrating directly with banks and financial institutions. They even planned to support some of Japan’s more unique payment methods until surprises during development made them change course.
Stripe’s entry into the Japanese market is both an essential case study for any FinTech company considering coming into Japan and an entertaining story for those of us with an interest in business in Japan.
It’s a great discussion, and I think you’ll enjoy it.
Podcast: Play in new window | Download | Embed
Subscribe: Google Podcasts | RSS | More
Links & Resources
- Check out Daniel’s blog
- Follow him on twitter @danielshi
- Find out more about Stripe
Transcript
Welcome to Disrupting Japan, straight talk from the CEOs breaking into Japan’s. I’m Tim Romero and thanks for listening.
Stripe is one of the largest credit card payment processing companies in the world and their Japan market entry did not go according to plan. It went well, mind you, but it just did not go according to plan. Stripe was agile enough to take the changes and surprises in stride.
Today, we sit down with Daniel Heffernan, the Japan head of Stripe, and he walks us through the process where one of the most technically sophisticated and streamlined fintech companies in the world came face-to-face with a very low tech and manual nature of fintech in Japan, and he explains how they made it all work. From detailed, extensive technical specifications that were delivered as uncopiable, printed documents in three-ring binders, to the Japanese aversion to interacting directly with banks and financial institutions, to trying to support some of Japan’s more unique payments, and some of the surprised they discovered once they began work. Stripe’s entry into the Japanese market is both an essential case study, for any fintech company looking at Japan, and an entertaining story for those of us with an interest in business in Japan.
But you know, Daniel tells that story much better than I can. So let’s hear from out sponsor and get right to the interview.
Tim: I’m sitting here with Daniel Heffernan of Stripe and we’re going to talk about Stripe’s market entry into Japan. And you guys have just officially launched officially but let’s back it up and talk about when you first came in. What was Stripe’s main motivation of coming into Japan in the first place?
Daniel: Well, when we started looking at Japan, we looked at it kind of like we do every other market that we considered. There are a few things we look at when we’re trying to decide whether to go into a market. One of them is the size of the e-commerce economy. Japan is pretty big. Last year it was about $130 billion, which is significant. That’s actually number 4 in the world. So you have China and U.S., are giants at the top, then it’s kind of a big jump down, and you have the U.K., and Japan is actually just behind the U.K. If you think about it from a population point of view, it’s really weird because the population of U.K. is like half of Japan.
Tim: Yeah, I find that surprising from both a population and an economy point of view.
Daniel: Right. If you think about why that is, it’s because of the number of online transactions that are happening online, or aren’t. So in Japan, I think it was 4.7% of all transactions are online, which is really small. As someone who uses the internet 4.7% feels tiny.
Tim: Yeah, so what is a comparable number in the U.S. for example?
Daniel: The U.S. is actually pretty much in line, but if you look at the U.K., it’s jumping over 10% and it’s up around 14%. So in the U.K., lots of transactions are happening online, even though the absolute value of commerce is smaller. So there’s this sort of gap between Japan and the U.K. in how much is happening online. And there’s an even bigger gap within this generation of our expectation how of much transactions should be happening online and what it actually is.
Tim: So Japan is still very much a cash-based society?
Daniel: Yeah, it’s cash-based, it’s offline. People aren’t buying things on the internet. These numbers include cash on delivery and things like that, so it’s transactions which are happening in the supermarkets instead of online shopping platforms.
Tim: Okay. So I take it Japan was not Stripe’s first overseas market?
Daniel: No. We look at the amount of e-commerce that’s happening online as one thing and it’s number 4, so it’s like, what is the 4th market you’re in. If you look at China, which is way up there at the top, it’s not a market we’ve gotten into yet. So another important thing is complexity and Japan is relatively complex. If it wasn’t, I don’t think you would have this podcast. There wouldn’t be enough to talk about but I think there’s plenty to think about when you’re thinking about trying to get into Japan and that’s because of all this complexity. So China is incredibly complex and Japan is pretty complex.
Tim: All right. So when you first came into the market, Stripe ran in stealth here for a good year, year-and-a-half or so?
Daniel: Yeah, we first got moving on Japan probably around January 2014. That was I was on the scene before, even, but we had identified SMCC, Sumitomo Mitsui Card Company, as a partner. We knew we were going to make it happen. We started opening the box and looking at what the technical integration might look like, what the deal would look like, what kind of agreement it is, but we haven’t started executing on it. So when I joined, which was April 2014, three months in, the first thing to do was to figure out this contract and then make a local entity so that we could sign the contract, and then we had a whole bunch of local vendors we would have to work with on things like data centers and things like credit card authorization switching networks.
Tim: You finally officially launched in September 2016, was it?
Daniel: October 4th.
Tim: October 4th, 2016. So in that two-and-a-half years, was that the expectation going in, or what took so long?
Daniel: So Japan actually wasn’t a long time for us. We have other markets where we’ve been in beta for longer. I think we approach new markets as engineers, so whenever we come into a new market—and I think this was true in Japan as well—we are working towards launch, and our partners are saying to us, “When is the launch? When is all the PR going to happen? When are you going to do the big reveal? When is the launch party?” And for us, that’s not something that we’re thinking about from day one. The first thing we should think about is getting to first charge. So the first live payment transacted in the local country. So it’s sort of a step by step process. We get to first charge, we bring in the first user, we run a private beta, and we just use invite-only, and we pick who comes into that. After that, we sort of move towards a public beta, where anyone can sign up, and for that, we need to be a bit more confident about the stability of the system, our risk in underwriting and KYC, (know your customer), ID verification plots. Once we have everything in place with the product market fit, then we move towards launch, so it’s a very careful step by step deploy. It’s not like doing this huge fork in your code, having this enormous branch, hitting deploy and praying that it will work.
Tim: Okay. Payments are particularly local. Every single market seems to have its own set of payment processing companies. And when you came into Japan, where there specific regulatory licenses you needed to get and how did you go about that?
Daniel: On the regulatory side, there are actually no specific relation yet, which covers credit card payment processors, and there is regulation coming, and it’s something where we’re sort of chatting with the Ministry of Economy Trade and Industry, making sure we’re in the loop there, we can see these things coming, and it’s interesting when you talk about payment methods. When we started working on Japan, the top of our list were things like Konbini, the convenience store payments where you, when you’re buying something online, you say you want to pay for the convenience store, they give you a number, or they give you a piece of paper to print out, and you go to the convenience store and you key it into the kiosk, or you scan the bar code, and you give them cash. Then the convenience store will tell the online seller that the cash has been accepted. JCB is a large Japanese credit card brand. And it’s not the largest in Japan anymore these days, but it’s still significant. And we were looking at these as sort of critical payment methods which we needed to have. We launched without them and what we realized while we were verifying that our product market fit was strong, was that these kinds of methods are critical for the businesses which are already in existence. I mentioned $130 billion a year of e-commerce in Japan. That is obviously generated by companies who already exist and I think this gap between this 4.7% and our expectation, for me, it’s like somewhere between 20 and 60% of transactions should be happening online. This is not going to be from these offline old companies coming online, and it’s not going to be from the existing online companies growing incredibly fast, or at least not all of it is. Those will contribute, but we think it’s going to be very small companies who are going to become very big companies, who are going to be generating. So we’re looking for fast-moving companies, who are looking at new business models, who are looking at targeting new types of customers, and who are looking to give new payment types to users.
Tim: Your initial plan was to support, let’s call them alternate payment methods, like convenience stores—did you end up supporting that?
Daniel: No. We did intend on doing that and I know we didn’t end up supporting it. If you look at the old existing companies, with basic e-commerce online, it’s about number of payment methods. They have a screen full of options you can click on and if you’re targeting those companies, you need to support all of those, or else you’ll be leaving some portion of volume behind. If you’re looking at these fast growing companies, they’re generally not looking at the same business model, or the same audience. We see companies who are platforms, for example. There are companies on Stripe in Japan, Drecom, which is a big gaming company, which is diversifying; they started focusing on sharing economy.
Tim: So was the decision not to support these alternative payment methods, was that based on the feeling that the market is moving away from them or is it based on a cost performance analysis where there wasn’t that much demand for them?
Daniel: It’s based on the idea that the businesses, which are going to be generating the future e-commerce volume in Japan are not going to be relying on these older payment methods. These companies need support for their platforms, they need payouts, they ID verification, it’s a more complex cycle, or they want to sell abroad, they need to have multi-currency presentment, they need to be able to have FX facilities. These are not things that come as standard in Japan, but the next generation wants these.
Tim: Once you got on the ground in Japan, what convinced you that was the case?
Daniel: When I started, or when we all came on board to Japan, we didn’t have customers. At first, we started off the idea of a partner who kind of knew who the partner was, we had to build out that relationship, the contract, the technical back end in processing, and we weren’t dealing with customers. But we got our first customer on board when we started our private beta in May last year. Once we started getting customers on board, initially we only supported Visa and MasterCard, then the customer, we had to learn about what it is that people want to do. One of our earliest users was actually a crowdfunding platform for record printing, old vinyl records, and it’s a very complex business model, but this is one of our first users, and they have this pay in and there’s payout, project owners, and project backers, then you hold the funds for a certain amount of time. And after you’ve raised enough orders to fund your batch of pressing, then they want to order more stock, which they use to sell separately. And it’s a very complex business, and this was one of the first users. So we very quickly understood that these are global businesses, these are platform business who, first of all, are drawn to us because of what we’ve seen as a trend abroad and what we’ve catered to, and second of all, we feel confident, or we at least are betting that globalization is going to happen, it’s going to be companies which are across border, which are global platforms who are the future of e-commerce. Not inward-looking domestic. Simple e-commerce.
Tim: So the original market was to support everything and you obviously trimmed that way back. So were there any specific features or specific changes you had to make to the product to make it work in Japan, other than the obvious integrations?
Daniel: There were a bunch of small things we had to do. For example, statement descriptors in Japan, shown in Japanese, they are shown in Kanji or Kana, the credit card issuer decides which one is displayed. So we have to modify our API to be able to accept Kanji and Kana and English. Really small nitty gritty changes we had to make but our general approach is not to fork Stripe in each country, but we try to make a global platform which works excellently in every other market. So when we identify specific uniqueness in markets that we need to support, if they are important enough, we try to make those something which is fed back into the global product.
Tim: That sounds like, from an engineering point of view, that was a pretty smooth path. What was your fundamental go-to-market? How did you acquire your initial beta customers?
Daniel: We’re very fortunate to have a very long waiting list of users in every country that we go into. If you go to http://stripe.com/global, which is the page where we direct people when we don’t support their country, we allow you to sign up and provide us with your e-mail address.
Tim: Were your original core users mostly foreign-owned companies or were they Japanese?
Daniel: I would say initially probably half and half. And gradually, that has become much more focused on Japanese companies.
Tim: As your presence here has become more know. Okay. That’s a great thing to have when coming into the market. For the first two years, you were running in, not stealth mode—you were running in beta—so you weren’t aggressively trying to market yourselves and expand your business, but what were you doing other than the engineering? How big was the team, or was it simply engineering for those two years?
Daniel: The first year was engineering. The first year was, I would think, maybe half engineering and half partner, legal type work, getting agreements in place to support the engineering. In Japan, even to get access to technical specifications, you’ll often need to sign documents and have agreements in place. They will be paper specifications, which you get in a big binder, each one of them watermarked. So there’s a lot of groundwork you’ve got to do before you can even get to the engineering work, so that was a big part of it.
Tim: And this was just the work to integrate with—
Daniel: First charge. To get the first charge.
Tim: Specifically, what types of integrations were giving you these big paper references?
Daniel: Well, there’s a couple parts that we have to build out always. One is the pay in, so there is the actually credit card processing. In Japan, there are two large credit card authorization switching networks. In a lot of countries, you can go directly to Visa or MasterCard, or you can go directly to a bank, and they will set you up, but in Japan, there’s this central switching idea. CAFIS is one which we operate on, and there’s a big spec that comes with that. And then we also plug directly into banks for the pay out part, so when we’re expense sending transfers to our users, we need to connect to a bank programmatically.
Tim: And that’s the ZENGIN network?
Daniel: That’s the ZENGIN network, exactly.
Tim: So it took a year of integration of reading, often incorrect, Japanese specifications, to get to first charge?
Daniel: Yes. And we’d spoken to lots of companies, especially I think foreign companies, entering Japan who have done similar integrations before. I think a key part of our valued proposition is we can come in and we can spend a year doing this, so you can get it done in like a week or two. It’s all behind the same API that everyone else uses. If you’ve used the API in the U.S., you may want to add a couple of parameters for the Kanji and Kana statement descriptors, but that’s pretty much it. The thing will work as it’s worked for you so far.
Tim: I would say that sparing users from interacting with CAFIS and ZENGIN is quite a selling point. So how big was the team there? Was the bulk of the engineering work done by the engineering team here in Japan, or was it done back at headquarters?
Daniel: We have our ops team, which does the sort of integrations back in San Francisco. There is a great team, and because we have a lot of integrations in a lot of different countries, they’re able to review each other’s code, they’re able to compare notes, they’re able to share knowledge, so we’ve decided, very explicitly to focus on having the team who builds these technical integrations in one place. So the hiring locally was primarily based on our support team, initially starting with a support manager, and the second pillar that we started to build, which was much later, was the growth team, which was in anticipation of launch. During the first year, just as we launched our private beta, we got our support manager in place, and she started to learn about how to translate Stripe into Japanese, how to communicate with customers, and I was winging it myself, sometimes in Japanese with our users until that point, but then we had a person who really understood Japanese and who knew what she was doing. And then just as we were getting into our full launch this year, we got our growth lead in place to look after any inquiry or any sort of sales leads that we get. So these are the two pillars we will be building the business on.
Tim: If the development was being done in the States, and all the specifications were in Japanese, that must have caused some kind of friction.
Daniel: Yes. Fortunately, I have a bit of a technical background, so I was able to do some basic back and forth there, and I didn’t have too much trouble or making sure that code would line up with the specifications that we had, but the first step is always get it translated. We would translate the entire thing and the translation would not be of perfect quality and it’s also difficult to make sure that the terminology is perfectly consistent. But we were able to get translations of the primary specifications that we had to work against. I would sit down with the engineers first and sort of walk them through the general overview, based on my understanding of having read the start of the specifications in both languages, and they would just go at it then, and I would be able to jump in and make sure that anything I was concerned about from the Japanese, was translating across into the implementation correctly.
Tim: So for that first year, you were very much part of the engineering team.
Daniel: I wasn’t writing very much code, but I was working every day with the engineers. And the certification is when it got interesting. Before you can go live, you have to go past the certification process on any of these systems, and generally, I had Skype open on my computer, to the engineer in San Francisco, and I had the partner on the phone. I was having both of these going at the same time and the partner would say, “Run tests 1-4,” I would say it in English to the engineer over Skype, they would run tests 1-4, I would ask if they received the tests, they would tell me if they worked or didn’t, and we would kind of go through this process.
Tim: So after you got to your first charge, what happens then? You ran in beta for about a year-and-a-half, so what was your focus then, during the private beta?
Daniel: There were a lot of things which were missing and which we had to build out, so for example, the site was in English until just before our launch, so there was a lot of work getting the site into Japanese and getting the code base in a place where it would be able to swap out strings at all, initially, since Japanese was our first major translation.
Tim: Oh, it was?
Daniel: France and Japan came along—our launches were aligned quite closely, so Japan was sort of the first use case we had in mind, and then we were able to get France along with Japan. I think we actually launched France a couple of weeks earlier than Japan but Japan was the initial driver of our need for sort of site localization.
Tim: Okay. So during beta, you were obviously fixing bugs, filling out the new requirements, but also ramping up support and marketing?
Daniel: Support primarily, yeah. A lot of time spent recruiting, a lot of time spent working with potential users, potential partners, people writing into us. Initially, all types of people are writing into us. We had a lot of people waiting for Stipe, so we had a lot of people who had gone to the U.S., Japanese companies whose CEO had gone on a plane to the U.S. years before, because they wanted to use Stripe, had set up a U.S. entity, had gotten a bank account in the U.S. over a few weeks, and then had set up an account and they just wanted to learn more.
Tim: Well, that’s a dedicated customer isn’t it?
Daniel: We’re very lucky to have a solid handful of users who have actually done that.
Tim: How big was your private beta?
Daniel: There’s no numbers that we make public on the size of our betas, but I believe at the end of the beta, we had everyone signed up. Not everyone was transacting every day, but we had thousands of people who were in the waiting list and who set up accounts.
Tim: All right. So now that you’ve launched officially, you’re obviously not going to be relying on the global brand and the English language advertising, so how are you going to be driving demand going forward?
Daniel: That’s a great question. I think all of the assumptions that we have about how this works in new markets basically don’t apply to Japan. One thing I think we’re going to double down on very early in Japan, and we actually just started hiring for this, is a coms role, which we don’t typically put into a market so early, but in Japan, I think brand is a very important driver of growth. I think the perception of a company very directly affects how well you can grow. People want to work with companies they can trust. That trust is driven by brand and driven by the perception that everyone is using, this new service or this product. It has to have that mix of feeling new, and innovative, and exciting, but at the same time, it has to be perceived as a major thing which everyone is using and if it’s too new and out there, then people will not feel comfortable using it.
Tim: It will feel risky.
Daniel: Risky. Risk aversion is a great topic.
Tim: So all of your users, to this point and going forward, have been acquired directly by Stripe? You’re not using a channel strategy at all?
Daniel: That’s right. In every market we’ve gone into, we’ve done direct sales. We obviously do have a lot of discussions with potential partners, and there are some relationships where we have large platforms using Stripe, which are so large that you could perceive them as being channel sales. We shall provide payments, for example, but it wouldn’t go to a payment service provider and have them resell Stripe, and go around with our marketing material kind of thing, which is a very easy thing to do with software. But with payments, it’s a little bit different because I think if you were selling software, you usually have a very large margin, and you can say, “Okay, we’ll give away however many tens of percent of this margin for someone who can get us sales.” With payments, it’s a very small margin, plus the usual incentive that you have is a percentage of that margin, and that percentage is usually measures as a percentage of the entire volume. So you could say, “We’ll give you 0.1% or 0.2% of the absolute value,” but if a user is getting very big, their absolute volume is becoming very large, they’re negotiating their rates, the rates are coming down, and to maintain this 0.-whatever percent to the reseller is a huge burden for us. And for the reseller, it can feel quite small as well. So even though we are selling software, when it comes to payments, which is a very low volume business, it introduces a lot of complexity into how you can get a reseller program up and running, and successful.
Tim: Yeah, I could see how it—yeah, just the economics just don’t allow for a lot of partnerships and distributor sales.
Daniel: We do work with a lot of dev shops. In Japan, you have the system integrators, the SI’ers. Japan is an interesting country, from the point of view that a lot of businesses don’t have their own in-house engineering teams. They’ll rely on these outsourcers and integrators. A lot of the time, the decision making around what kind of tooling or what kind of systems are used is actually coming from the system integrators, as opposed to coming from the actual businesses themselves.
Tim: People are shocked how shallow IT knowledge is here.
Daniel: A lot of the time, it’s literally an IT department with a person who will throw up some requirements, throw it to a few vendors, who may throw it to some more subcontractors. There’s not actually much engineering work going on at all in the company, so in that case, it’s important for us to talk to the business and to understand their own requirements as well but to also work with the system integrators so they understand what the value proposition of Stripe is, how it can help any of their clients.
Tim: So are the systems integrators driving a lot of your sales in Japan now?
Daniel: Not a lot of our sales, but I think we see it as a very interesting way of growing. It’s actually a separate discussion to a reseller model. You don’t have to actually see it as channel sales because what we’ve seen globally and in Japan, with a lot of the companies we’ve worked with so far, is it’s actually a clear win-win. They’re getting better technology, they have tooling available that their competitors don’t have, the integration time is quicker, the quality of support that we give them is potentially higher than what we get from other companies. So even though there is maybe not a direct financial incentive, a kickback of some description, there are a lot of other incentives that are lined up for them. So we work with a lot of these companies in Japan. We have fantastic relationships. We get on really well with the system integrators, we kind of run some basic events and things together. We can do that without a reseller agreement.
Tim: Excellent. Now you mention that you think the future of online payments in Japan is going to be more towards credits cards, less towards COD and these convenience store payments. Do you see you competitors in the market adapting? Do you see them changing their behavior because of Stripe in the market now?
Daniel: I suspect there will be movement from the existing players to try and keep up and I don’t think that’s going to be necessarily because of Stripe, per se, but I think it’s going to be in response to the movement and the market. You’re right that we anticipate that cards will increase as a payment method. I think that’s primarily because if you look at the types of businesses that are happening, if it’s a complex platform, you need to be able to refund easily, for example, which you can’t do with cash on delivery, you can’t do with convenience, you can’t do with bank transfers. Or, for example, if it’s a platform where you have sellers and buyers, and you’re scalably achieving sellers, you could potentially have people on the platform who are somewhat fortunate, or who are not fulfilling the services to the quality that they need to be, so you need to have a dispute resolution process, which credit cards have kind of built in. There’s disputes, there’s chargebacks. And also, if you look at going across borders, and if you look at selling into other countries from Japan, credit cards are a really great way to do that.
Tim: It’s almost your only option, really.
Daniel: Exactly.
Tim: Two things you mentioned stuck out at me. One, I understand that credit card fraud is still relatively rare in Japan.
Daniel: Credit card fraud has traditionally been very low in Japan, and I don’t anticipate that credit card fraud from within Japan will increase, but I do expect a lot of Japanese companies are going to be selling more and more abroad, and as we approach the Olympics, this is a big thing. People talk about the Tokyo 2020 Olympics and about the inbound and about all these visitors coming to Japan, then buying things in advance of coming, or continuing buying things after they’ve left, gone home. I think this idea of cross-border trade, where China, the U.S., and other countries are buying from Japan, is only going to increase, and with that, there is going to be a lot of fraud.
Tim: Is it something that merchants are concerned about now in Japan?
Daniel: Yes. We have spoken to many large Japanese companies who have started selling abroad, they’ve been hit by fraud, and they’ve wrapped it up—they stopped, that was the solution. They didn’t have tooling to help them figure this out. A lot of them would maybe first, they would have one person who would review every order and say, “This one is above $500,” or “This one is coming from this country, we don’t want to take this,” and kind of go through these manually.
Tim: That’s not a very scalable process though.
Daniel: It’s not. It’s something you can do, and if one person is doing it, they can learn to identify a lot of these signals, but what you really need is scalable tooling which can make a lot of these decisions for you. And in Japan—in a lot of countries it is just starting to come on—but in Japan, because there traditionally hasn’t been a lot of fraud, there hasn’t been demand for these kind of products, so it hasn’t really existed. Just recently, we announced a product called Stripe Radar, which leverages machine learning to allow businesses anywhere in the world to leverage the learnings that our machine learning is able to glean from our global network of sellers. So we’re in countries like Mexico, where fraud is a bigger problem than Japan, and if there’s some new type of fraud that is happening in Mexico, the models can pick it up and immediately protect Japanese sellers from that kind of fraud.
Tim: It’s interesting that the Japanese merchants are becoming more and more aware of this problem.
Daniel: Yeah, I think they’re becoming very aware—or they’re already very aware—of the opportunity of cross-border trade. Many of them are becoming aware of the disappointment that can come from fraud once you start trying to do that. I think there’s a big pool of companies out there who are ready to sell abroad, they are strategically planning to sell abroad, and they’ve tried at some point over the past few years, and they’ve been bitten, and they had to block it off, and they’re not sure what to do next.
Tim: I would imagine that this global outlook and this fraud prevention would be a rather unique selling point for Stripe in Japan right now.
Daniel: Yeah, we get a lot of people asking us about our multi-currency, and then also about fraud prevention tooling.
Tim: Excellent. It certainly sounds like you guys are well on your way. The first hard thing is complete and you’ve launched out a beta, but on a personal note, kind of looking back, over the market entry, over the last two-and-a-half years, what would you say the biggest mistake you made was? If you could go back and do something again, what would you change, or did everything pretty much go as planned?
Daniel: I wouldn’t say that everything went as planned, but I would say that both me personally, and the local team, and Stripe, have started off with a very strong hand. I have a personal relationship with our co-founders, for example, and we’ve found some very great people to build out the local team. We have HQ, which is not doing a top-down, “Here is the strategy. Go and execute,” sort of style. It’s more based on decisions made by the local team. We know the market best and that’s he understanding throughout the organization, through all of our markers.
Tim: That is a very enviable situation to be in. Let’s talk about that. You worked with Stripe and you knew the co-founders for quite a while before you came to Japan, right?
Daniel: Right, yeah, we met when we were in school.
Tim: Did that relationship transfer over to the Japanese team as well, where if the team on the ground said we need these resources, or we need this time, headquarters accepted that?
Daniel: I don’t think anything so direct as I would ask for something and we could get it because there’s a personal relationship. We’re specifically trying not to become a political organization where everything needs to happen, but I think having a high trust relationship to start off on makes things very different. I understand that any decision that’s made by our exec team is made with everyone’s thoughts in mind. I implicitly trust decisions that are made by that team and I think the team implicitly trust the decisions that are coming out of the Japan team as well. Without even going towards specific asks of each other, I think just having this basis of trust allows us to not nit-pick each other, or jump into thinking that one knows better than the other. I think we all clearly know who is doing what, and respect that, and from what I’ve heard from some other people who have been running market entry for companies coming into Japan, it’s very rare to have such a high quality relationship with the exec team and HQ.
Tim: That is fantastic. What else do you do to maintain that? Is there a lot of travel? Do the team from headquarters come out here to Japan? The Japanese team go there?
Daniel: We do a bit of back and forth, so I go out to San Francisco about a week a quarter, which is not a lot, and I probably should do more, but it’s always very hard to take your eye off Japan. Every time I do that, I feel I should have done that earlier, I should do that more often, but—
Tim: Suddenly priorities change and things get done faster.
Daniel: I think even just keeping that face-time is incredibly important.
Tim: Yeah. There’s Skype and GoToMeeting are wonderful, but there are things that only get done when you’re sitting across the table from someone.
Daniel: Absolutely. And we’ve been very fortunate to have people visit from the U.S. quite a lot. We’ve had a lot of our management come out here to help us, and just on vacation. And while they’re in town, we can meet up with them, we can talk to them a little bit, and I think Japan is probably an easier country to do that from. It’s quite interesting, and it’s also relatively close to San Francisco. It’s a little bit further than Europe, it’s further than Australia, it’s further than a lot of Southeast Asian countries.
Tim: It’s great to be managing a Japan team that has that degree of independence. I think it increases the chance of success tremendously. But have there been any parts where, for example, a different sales culture has developed in Japan, than in headquarters? Has a local culture also developed with this independence?
Daniel: I think we absolutely do have a local culture, but I’m also cautious of having anything too unique in the local market. I think there are certain changes or certain allocations which are very important but we have to be very careful that we stay as a global company. When people join Stripe Japan, we don’t talk about them joining Stripe Japan; we talk about them joining Stripe. I think the contracting entity who is employing you is a minor detail.
Tim: What’s an example of something that would be unique to Japan, and important, and something that you would want to remain as part of a global culture?
Daniel: One thing that we do, that I don’t know we would do in other markets, when users write into us in Japan, a lot of the time, they will, in their first e-mail, say, “Can we meet?” They may not even tell us why they want to meet but they really require that face-time, and in other markets it’s more transactional. And we’re not really set up to do that. There’s no culture, really, of meeting people in face so much. In smaller markets, like in Australia, or Europe, it’s something which can happen, but if you look at our larger markets, like the U.S., it’s not something that happens all that often. So even if it’s someone who is writing in for the first time, we will ask them why they want to meet and ask for some details, but we’re very much interested in meeting people, talking to people, and giving them what they need to feel assured that we are a real thing, that we have people here, that we have an office.
Tim: That is a difference in the Japanese sales cycle. That face-to-face meeting, even for relatively small transactions, is very important here.
Daniel: Yeah, and it’s all about trust, and it’s about avoiding risk, as you mentioned earlier on. Making sure people understand that you do really exist, and just for them to look into the eyes of the person on the other side, and to realize that there’s a person there and it’s someone they can trust.
Tim: Japanese consumers and businesses are notorious for demanding high levels of support. Have you had to adapt your support team in any way to meet that?
Daniel: Yes, we definitely maintain a higher SLA, or it’s an internal SLA, but we have a very high bar for our response times and our resolution times in Japan, compared to other markets. And it’s something which has resonated well. In every market, we try and go beyond expectations, which is relatively easy to do—or it’s not that easy, but you can go beyond expectations in support in other markets by responding within a day.
Tim: Yeah, in the U.S., the expectations are pretty low.
Daniel: Right. And you can go beyond those. And in Japan, the expectations are very high, so it’s a bit more work to keep ahead of that. We have had plenty of users come back to us and specifically tell us how happy they are with how quickly we have been able to get back to them, resolve their issues, and we definitely want to maintain that, but it’s definitely something that is a challenge to maintain, and we have to keep hiring, we have to keep the quality high, and especially in a market like Japan, where English levels are relatively low compared to a lot of markets. I think hiring scalably for a team like this, where you’re looking at teams and documentation coming from the U.S., and translating that to Japanese, and doing that without making mistakes, and fast and high quality, is going to be a challenge that we’re going to see.
Tim: Do you see any of these Japan best practices filtering back to headquarters?
Daniel: Yes. We are working very closely with our user ops team globally, which is our support organization over there and we’re also feeding back some learnings from the growth side as well. And those learnings wouldn’t be things like you should have lots of face-to-face meetings, but the partnerships, the relationships we have with these shops and SI’ers in Japan is something which is very interesting, and it’s something we can eventually scale to other countries. So it’s something we’re looking at and we’re feeding back up. Those are our two recent examples, but we’re very much about feeding these learning through and getting them into the organization at large.
Tim: Most country managers, and certainly all country managers that come from headquarters, this was you first time doing this. On a personal level, what was your biggest challenge personally in the last two-and-a-half years in pulling all of this together?
Daniel: My background is as an engineer. When I initially talking to people at Stripe about joining Stripe, it was when I was finishing my post-grad in university in Japan. I had never worked in a real job as a full-time employee before. I remember specifically saying, there’s no way I could make any of this succeed because I have no idea what I’m doing. I had never done any kind of sales work, I had never done any kind of BD, I had never looked at a legal contract in Japanese before, never mind trying to negotiate one in English. PR, marketing, you name it—I had zero experience in it. I did go into a Japanese company, I was working at Cookpad for a year, which was a fantastic experience for me, but just having that one year experience made me feel a little bit more confident in trying to do something at all in Stripe, but I was coming from a background of post graduate in Creative Informatics, which was kind of a natural language processing for me, and a year as an engineer in Cookpad, and that’s not exactly the right background for someone who’s going to spin up one of these.
Tim: So I guess almost everything was challenging for you then.
Daniel: The engineering part wasn’t too bad. Everything else was a new challenge for me. So I’m learning slowly but surely, but the engineering part is still just about the only part I have confidence in.
Tim: Well, there’s something to be said about approaching a problem with fresh eyes.
Daniel: I think there’s a story our co-founders tell about when they had this idea for Stripe, and there should be the same API, which allows you to send payment information and gives you money. It’s pretty straight forward. We have things like AWS, where you can hit an API and you get a server. One looked at the other and said, “How hard could it be?”
Tim: That’s always a dangerous question.
Daniel: So I think that naivety is a good thing but maybe best taken in measures.
Tim: Okay. What would be your best advice to someone else coming into your situation, someone else who has to bring a technology company into Japan?
Daniel: I think the best advice that I could give is to know your users. As I spoke about earlier on, the plan had changed significantly, based on the input from our users. I think we were very quick to throw away any assumptions that we had if we saw evidence from our users that our thinking wasn’t correct. Also, as I mentioned earlier on, we don’t come up with a plan, run the plan very quickly, and say, “Here it is. That’s done.” It’s an iterative approach, which allows us time to get to know our users, allows us time to bring in their feedback, adjust the road map.
Tim: So listen to your users in Japan and adjust the plan as necessary?
Daniel: Yes, which is something I was fortunate enough to have an organization which supported that approach. That may not be the case for everyone who is coming into Japan, but if you can convince the company back home that you don’t know Japan—which is, you’ve probably been hired because you know Japan—but you, as a company, don’t know Japan, and you won’t know Japan until you really hit the ground, you start things moving and you can learn about the market as you go forward. I’ve seen a lot of companies come in with assumptions and they’ve seemed to all have been wrong.
Tim: It seems to be a lot of thinking like a startup again.
Daniel: Yes. That’s a very fair thing to say. I think, in a way, if you can get the organization in the right place, it’s kind of like a really imbalanced startup. You have a product already, you have a team somewhere—maybe in the U.S.—you presumably have customers already in other countries, you have revenue streams, you have funding, and now you’re bringing this to a new market. So you kind of get all the cool, fun parts of a startup, of trying to find out what the customers’ needs are, trying to find a product that fits that perfectly, but you have the bonus of having a product that already exists, which can be adapted, and you’re not starting from scratch. I see it as quite an unfair job, in a positive way, to be able to produce something, which has such a high impact on a local market, which is really changing business in Japan. It’s enabling news types of businesses, which is something I’ve always wanted to do. It doesn’t require you to go and find funding, and for you to throw away your product and start rebuilding it because of the pivot.
Tim: It’s just the fun parts. It’s just that product market fit, that business development. That’s a great way of looking at it. Now that you have expanded your user base, and you’re moving beyond the initial users who were just fans of Stripe to begin with, do you see changes in your positioning, or do you see changes in the product because you’re listening to this new, larger set of users?
Daniel: So by listening to our initial users, who are inherently looking towards Silicon Valley, they’re inherently trying to find out what the most recent business models, the newest, most cutting edge business models are, and how those are being implemented by people who are running these companies. That feeds into our product, which is not just from Japan. We’re working with cutting edge companies in Silicon Valley. They’re asking us questions and trying to get us to build things for them, which we’re building. And when we bring that to Japan, we’re a few steps ahead of the market. I think the startup market in Japan is objectively behind the U.S., depending on how you measure it. It’s anywhere between 40 to 75 times smaller than Silicon Valley. I think being able to bring a product to the market—you have companies in Japan who are building, for example, the next Lyft, or the next Instacart, and they look to the U.S. and they say, “What are these companies using? This is a business model that should work great in Japan. I know how to execute this,” and they see that it’s using Stripe and they say, “Okay, if we use Stripe, we will be able to bring this technology to the Japanese market as well.” So they are doing market entry as well. They are looking at bringing these business models to Japan, adapting them to the local market, and we’re providing a common platform for them to move on.
Tim: Okay, so I guess from headquarters point of view, this makes a lot of sense. Listening to your users isn’t listening to your Japanese users, or listening to your French users. It’s listening to all of your users, everywhere.
Daniel: Yes, exactly. And the result of that, which is pretty cool, is you end up with a product that is ahead of the curve. It’s a few years beyond. I mean, we’re, with for example, Relay, or with Atlas, which is some products we have. They are a little bit ahead of the curve. Atlas allows you to set up a U.S. entity, to get a U.S. bank account, all through a web form, and we do this all for you. We’ll give you a Stripe account, tying these things together, so if you’re an entrepreneur in Egypt, or in Cambodia, or wherever you like, and you need to get up and running, and you can’t afford a ticket to get into the U.S., and you don’t have local infrastructure, we can get this approval. This is quite ahead of the curve. But we’re building these tools, which are products which we’ve learned about the need for from our users who are trying to push the envelope. We build them and we quickly find that there’s actually quite a large market for these kinds of products. So in Japan, even, one of our launch partners was ANA, which is a relatively old, traditional Japanese company, it’s the largest airline in Japan. People may not perceive it as a tech company, but we see a lot of these large companies who are trying to innovate. Even in the U.S., we have Walmart, we have salesforce. We have traditional companies who are looking at the smaller, upcoming companies and saying, “What are they doing? How can we innovate as well?” We’re fortunate enough to have relationships with those companies. So with ANA, we’re doing a crowd funding platform. In order to pull off a crowd funding platform, they needed a piece of infrastructure that could help them with that. They knew that we powered Kickstarter, that we powered Indiegogo, and that we were available to help them in Japan as well. By working with these companies like Indiegogo, these people who are really trying to do new business models, building this technology, we’re able to provide that to not just these companies, but there’s a huge market of small and large companies who need innovative infrastructure.
Tim: Excellent. So thanks for sitting down with me. I really appreciate it.
Daniel: Thanks so much. It’s great.
And we’re back.
I thought Stripe’s decision to have the customization for the Japanese market done at headquarters was an interesting one. This is normally a recipe for disaster, but I think there’s a two-fold reason that Stripe pulled it off. First, the team at HQ was dedicated to developing localization and market-specific functions. They were not on loan from the so called “more important” core development team. These types of integrations, where the first priority, and the usual struggle for resources was avoided. Second, in Daniel, Stripe had a highly technical, thoroughly bilingual engineer to coordinate their efforts. I’m not sure either Daniel or Stripe really appreciate how unusual that is here in Japan.
Now, one of the most interesting things about Stripes Japan market entry strategy was the decision not to support popular payment methods, like COD and convenience store payments. We’ll give them the benefit of the doubt and assume that they were not just rationalizing a decision not to do a lot of development work that would only apply to the Japanese market because Daniel made a very good point during the interview. E-commerce transactions are becoming more and more complex.
In fact, the extent to which online business models can support refunds, and escrow, and ongoing payments, and digital goods, and a whole host of other activities, mean that they are far more flexible than cash, and certainly more flexible than COD or many of Japan’s touch based e-commerce solutions. I think Stripe’s bet that these new business models will drive e-commerce in the future is a good one. Traditionally, commerce has been mainly the simple exchange of cash for goods, but the internet and flexible payment systems like Stripe open up the field to a far more complex and innovative set of transactions, and hopefully we’ll be seeing a lot more of them in years to come.
If you’ve got a question about Stripe or their market entry into Japan, Daniel and I would love to hear from you. So come by DisruptingJapan.com/Show066 and let’s talk about it. When you drop by, you’ll see all the links and sites that Daniel and I talked about and much, much more in the resources section of the post.
And hey, I know you’ve been meaning to do this for a while now, but when you get the chance, please leave us an honest review on iTunes. It’s really the best way you can support the show and help us get the word out.
But most of all, thanks for listening, and thank you for letting people interested in Japanese start-ups know about the show.
I’m Tim Romero and thanks for listening to Disrupting Japan.
Loved the show. Been using Stripe in the US for Second Harvest Asia and loved how efficient payments are. Just opened an account for Second Harvest Japan and look forward to working with Strip Japan. Two questions: 1) I could not find information on charges for clearing payments. One of my beefs with credit card companies is that they make money on charitable donations (particularly in natural disasters) because they charge a percentage of the donation. I understand part of the reason is insurance. 2) I think Strip should reconsider the market of making payments via convenience store. I think there is a large market there in which people want to pay without having to tie up a credit card.
Would love to meet some time to see what we can do together.
Good luck and great show.
Hi,
Thanks for listening.
I think Stripe’s decision not to support connivence store payments was an interesting one. They are clearly giving up some market in the short run since this is a very popular payment method in Japan. However, they are banking on the fact that future growth will come from more complex transaction models that can only be facilitated with credit cards. We’ll have to wait an see how it plays out, but there is a good chance it will work out for them.
Tim