Finterview: Airbnb’s Jonathan Golden
Jonathan Golden joined Airbnb six years ago as the company’s first product manager and has since helped build key parts of the platform including international and host products. Today he is Director of Product leading payments, monetization, and business travel. William Hockey sat down with Golden to talk about building payments infrastructure, the challenges of operating an international marketplace, and advice for disrupting competitive industries.
WH: What was Airbnb’s payment stack like when you joined in 2011?
JG: The payments stack at Airbnb was quite innovative at that time because the payment industry’s technology for marketplace businesses, where you facilitate the transfer of money between two sides, was not robust. The industry only focused on basic e-commerce. Airbnb leveraged Braintree and PayPal for payment processing and one bank to send money to hosts. We processed in one currency, which was USD. Many elements were manual across the board.
As the business grew, payments had to grow with it and scale, and so we had to continually build on that infrastructure. Payments was starting to fall behind the business, so when I took on the role of doing a lot of international work, I also started to do a lot of payment expansion work into those new markets.
WH: Airbnb has almost turned into a payments company, but it didn’t necessarily start out that way. What was the pivotal moment when you realized that, in order to be really successful and scale, you needed to become a payments company?
JG: At least for me, it’s when we went international, and we realized how important payments were to facilitating human interactions. We needed to create trust between individuals and Airbnb handling the flow of funds was paramount. When hosts have a confirmed booking, it’s important that they have confidence that they’re going to get paid. It’s great for guests to know that they can get their money back if they have to change their reservation or if the accommodation isn’t what they expected. Through controlling the flow of funds, we could resolve any potential issues between guests and hosts, creating a platform of trust.
Years ago, the money piece of booking online was still handled completely offline. It was uncertain and cumbersome. But we realized that when you move that transaction online, it enhances the real life interaction between people. Money is no longer a concern for them –we’ve removed it from the equation. And that’s what we’re really going for. The mission of the company is to belong anywhere, and that requires people understanding each other and not letting other things get in the way. That’s the top priority for the payments team; ironically, by us handling the transaction seamlessly, we make the experience between guests and hosts less transactional and more human.
WH: Can you go into what the payments infrastructure actually looks like?
JG: Yeah, absolutely. We work with roughly 20 different vendors across the board, creating a global payments network. I would argue that we have one of the largest global payment footprints in the world. We process payments in more than 35 different currencies and operate in 191+ countries. Our most extreme example recently is Cuba. One of those 20 vendors we work with is handling Cuba alone since it’s such a unique and complex market.
WH: I recently went to Cuba, and I used Airbnb, and it was such a seamless process. But when I got there I couldn’t use my credit card anywhere. So how did you get that to work?
JG: We found a partner who was able to actually hand-deliver CUCs, the Cuban currency, by putting people on bicycles and dispatching money on a daily basis. The partner also has an entity in the United States so we’re able to facilitate that transfer.
In terms of the payment stack, though, I can give you a little bit more details.
WH: Yeah, I’d love that.
JG: Because of the nature of Airbnb’s global reach, we’re not just transferring money within a particular country; we are also taking that money and transferring it elsewhere. So when you travel from say France to Brazil, we’re taking Euro from France via a credit card and finding a way to pay it out directly to the host in Brazilian Real.
For other companies, it’s much more siloed. If you think about an eBay-type model in a particular country, most of that money is going to stay within the borders. But for us, it’s the exact opposite: Most money is transferred across borders.
We have several processors we work with to process credit cards through Braintree. We also leverage PayPal, Alipay in China, and several local payment solutions in Europe, South America, and other markets.
And what’s interesting is that most of our business is growing in markets where we don’t have local payment solutions. We have a certain percentage of funds that we always want to cover with local payment solutions to make sure we deliver a good experience. Over time, that always trends down, because, as we grow, we expand into frontier markets.
We have more vendors on the payout side, because it’s a less-developed part of the payment industry. We use several banks to facilitate payouts through ACH transfers in the United States, SWIFT transfers in Europe, or local methods in other countries. In the more frontier markets we use Western Union, which we found can reach a lot of countries where other payment solutions can be a little bit more challenging.
WH: As you move into new markets internationally, do you use a generic backend payment system developed here in San Francisco, or do you have local teams that are responsible for each area’s payment infrastructure?
JG: All of our engineering technology is based in San Francisco. Payments is a complex, cross-functional discipline. We have really strong support from operations, finance, legal, compliance, and business development. We do touch a lot of international offices, especially around entity structuring, contracting, and making sure that we follow regulations around the world.
On the technology side, payments is structured into four teams right now, which will evolve going forward.
Right now, the top level is called payment experience, which is managing the front-end user interaction for all touchpoints dealing with money. Below that, we have a team called billing platform, which is basically our internal API. Any other team that wants to implement any form of payment can work with that team to make it happen. Below that is the payments infrastructure layer, which deals with all of our processors and gateways. And then below that is our financial infrastructure layer, which really deals with reconciliation.
Tie those together with our reservation data, external processor data and external banking data, folding in all the different entities around the world—that’s a big endeavor. We’ve focused a lot of effort around tracking each and every transaction more recently to have a high level of fidelity and to understand of our flow of funds in order to make the product experience better.
WH: How do you make the build-versus-buy decision?
JG: For us, it always goes back to the guest and host. We want a certain level of experience for them, and if we feel that we can deliver that experience in a compelling way with a buy solution, and that buy solution is going to scale for us and not add a tremendous amount of complexity to our system, then we’d be happy to go with it.
More often than not, though, it tends to be more of a build solution, where we want to control the user’s experience much more delicately. It also helps with flexibility, whether it’s changes with the flow of funds, or the flexibility of when you pay, or when we issue a payout.
Other considerations, obviously, are transparency, where the money is, and the timeliness of funds. And, for payments, uptime and redundancy are huge for us, too.
WH: Payments in some companies is seen as a cost center that can consume engineering and operational resources without providing a new feature or converting users. How do you get engineers and team members excited about working on a backend payments problem? And how do you get leadership to focus on payments opportunities?
JG: Payments is core to pretty much any marketplace model, and our payments platform is, not surprisingly, core to Airbnb. When you unwrap the covers, and you start explaining this to potential new team members, they get excited.
In terms of explaining it to leadership, I think you start by recognizing that payments is one of the biggest friction points for any user in any flow. You can see that manifesting itself in the real world. For instance Amazon is fundamentally redesigning the grocery store focusing on the checkout experience and removing it in its Amazon Go pilot stores.
Payments is arguably the most critical piece of commerce whether it’s online or offline. If you can get that right, you can really enable a much better experience.
WH: You’ve been at Airbnb for six years. What have you done wrong in that time? What do you wish you had done differently?
JG: I think we focused a ton on scale and the ability to process internationally to enable the business growth. We were playing a lot of catch-up regarding the basic underlying architecture to build a platform that was more than just a reservation booking platform. So recently, we’ve rebuilt a lot of our infrastructure with micro-services to enable more flexibility and more use cases. This new flexibility allowed Airbnb to launch Trips in the fall where a guest can book an experience, not just an accommodation, while traveling. I wish we would have built that flexibility earlier.
WH: How did you navigate fairly aggressive competitors or incumbents, and what advice do you have for startups facing pressure from incumbents trying to shut them down or push them out of the market?
JG: I think every industry is competitive. In terms of Airbnb specifically, I know this sounds silly, but we really try to build a great experience and that trumps trying to mimic a competitor. A lot of competitors that came after us in the earlier days tried to solve a lot of problems in a much more manual way to get to scale extremely fast and didn’t really deliver a good product experience, especially when things didn’t work out like during a rebooking. So focusing on product is key.
Airbnb is inherently a global network effects business, which means that the utility of guests and hosts increases as more participate on the platform. That dynamic created an opportunity for us to distinguish ourselves because we scaled before others did.
I encourage entrepreneurs to try to find ways to create some form of viral loop, even if they don’t have inherent networks effects, to enhance growth so they can hit escape velocity before others.
WH: You were insightful enough to have joined Airbnb really early—it seems as though you could tell it was going to be a really big success before it actually was. How did you find out about Airbnb? What signals would you encourage others to look for when searching for a startup to join?
JG: I learned about Airbnb in 2009 from an entrepreneurial friend. I was immediately struck by the design caliber of the website. In fact, when I used to think about my own startup, I would start with their website and then iterate from there in terms of design. So I always had an affinity toward what they were creating from a product standpoint.
I tracked the company, and I started to realize that—with the advent of Facebook Connect bringing identities online and the ability to start transferring money online—there was a global opportunity for a business like this. I looked at a lot of different prospects when I wanted to join a startup. Airbnb had an extremely high level of risk because it was early and there was still a lot of uncertainty around the product and business. Airbnb created a totally different behavior, letting somebody else into your home. I asked myself, if this worked how big could it be? After I pondered this for a few days, I realized that though risky, if Airbnb could unlock this new behavior, the opportunity would be immense. I also loved the people, and I loved the sense of energy, and I loved that they were really focused on design, and I thought that design was going to be something really important going forward in the consumer space.
Those, to me, were the signals. And the thing about the team that was so exciting was that they were so committed and so up for trying anything to make it work.