How can a simple offering be transformed into its own platform? Renaud Visage, Co-Founder of Eventbrite, and Romain Huet, Head of Developer Relations at Stripe, know what it takes to effectively evolve your offering into a platform without losing what made offering appealing in the first place. In this session, Renaud and Romain use 5 simple steps to help evolve your offering into its own platform.
Want to see more content like this? Join us at SaaStr Annual 2020.
Renaud Visage | Co-Founder @ Eventbrite
Romain Huet | Head of Developer Relations @ Stripe
FULL TRANSCRIPT BELOW
Romain Huet: Good afternoon everyone. Thank you so much for being here. It’s a great pleasure to be here today with Renaud, Eventbrite’s technical founder.
Renaud Visage: Thank you.
Romain Huet: When you look at the most valuable companies in the world today, as I was just mentioning, they tend to be platform companies, when you think about the Facebook, Google, Microsoft, Apple, all of these, we tend to define them as platform companies. For a few companies like, including Stripe and Twilio, for instance, we stayed with the API in mind from day one, trying to think about the building blocks for developers to build upon. But most companies, they tend to start with an actual core product, and I think that’s what we’re going to talk about today.
Romain Huet: See how you can turn like a SaaS business, or a core product you have and make it a very successful developer platform. Maybe we could start by even defining what a platform is because it’s pretty hard. A lot of startups out there tend to define themselves as startups who aspire to become a platform, and it’s quite nebulous to define. Right, Renaud? For me, one definition I like is the one from Bill Gates when he said, “the platform is when the economic value of everybody that uses it exceeds the value of the company that creates it”. That’s what a platform according to Bill Gates. What would be your definition? What do you have in mind?
Renaud Visage: I think there are a lot of different types of platforms. So understanding the differences between all the ones that are defined somewhere, and you can find lists of 9 16 20 different types of platforms. But in the end, I think a platform is way for one product to connect with other products. In its most basic definition, I think that’s what we at Eventbrite benefited from the most. If you think about it that way, then I think it simplifies a bit the equation, and you can think of how does that apply to your business?
Romain Huet: Of course, besides the brand I mentioned earlier like, Google, Apple and Facebook, and so on, a lot of marketplaces also like Uber and Airbnb they tend to think of themselves as platform because they have the buyers and the sellers connecting to each other. I think on top of your definition, I would add that in addition to that co-creation of value for consumers, what I really like about platform is the fact that they’re like pieces of infrastructure, they’re building blocks for developers to build upon.
Romain Huet: I think that’s the mindset we’ve had at Stripe since day one. Like, how can we provide economic infrastructure for developers to build applications and services and in a weird way, future proof? Let’s start with our five key steps. Let’s start with the first one, which I think really is to prepare your core product for scale. Can you tell us a little bit about the early days of Eventbrite and how you kick started the product?
Renaud Visage: Yeah, I think before you can run, you have to learn how to walk. We started Eventbrite with mostly a tool for event organizers to be able to sell tickets very quickly. Any type of event organizers, so small events like this ones all the way to larger events. But we really focused on acquiring one side of the marketplace first, which is the event creator, because they are the ones who have the inventory, the connection with the attendees. We thought that if we could grow that side of the business, then we could extend into a platform down the line.
Renaud Visage: But the initial early years were, I would say 120% focused on creating that core product, acquiring these first customers that then we could embark into a platform offering, which came much later. But that was quite different for Stripe, because it started as a set of API’s, basically.
Romain Huet: Yeah, we started as a set of API’s. The idea for Stripe, I’m sure most of you know in the early days was to have just a few lines of code and lead developers accept payments in the apps and services. What I like to know always when I meet founders is how they managed to get their first users. How did they manage to bootstrap the company, because I’m sure many of you have read this essay from Paul Graham when he refers to what he called the, “Collison installation”. So Patrick and John Collison, our founders would spend a ton of time with the early Stripe users in actually helping them write the code to install Stripe. That’s how they got started. Did you have any tricks for Eventbrite to kick start the product?
Renaud Visage: No, I think we had a great network in San Francisco, of event organizers. TechCrunch was very much nearby, and they were the first ones to have fairly large events. They were one of our first customers. We worked really hard to understand the pain points. We didn’t come from the even business. One thing we understood very quickly was that access to money for organizing an event is quite critical. So the first product we launched had an integration with PayPal that made it very easy for the event organizers to get all their ticket sales directly into their PayPal account as they were happening in real time. That gave them the cash flow that they needed to power their events and organize and book the room and book the acts and all the talent that you need to make an event successful.
Renaud Visage: We spent a lot of time with this crowd. Tech events in San Francisco, then Tech events in New York, and Tech events in London and then started. I think we were lucky to have the first mover advantage. There were solutions for the stadiums for the very large conferences, but nothing really for under 20,000 people. That was really our core focus, and extending beyond Tech then we found a different set of requirements. We went to beer festivals, classes, concerts, theaters, et cetera. That added different requirements for each type of organizer that we incorporate into the product, which was not a platform at the time.
Romain Huet: Yeah. So when you build that product, and you saw it grow overtime with those first types of users, at which state of the company did you think about becoming a platform? Was that something you thought about from day one, but you pushed in the future or did that come along later on?
Renaud Visage: Well, we were built on top of another platform, PayPal, so we knew that API’s were going to be important for the future, that data silos were not going to be sustainable, because data wants to be free. I think that you’ve heard that before. We knew there was probably a platform play down the line, but we just had limited resources. When you have limited resources, I think that’s why step one is to build your core product, and acquire this mass of customers that you can then build for with a platform. So it was always in the back of our minds, just not a priority because we didn’t have the staff to power that.
Romain Huet: Right. Maybe that brings us to the second point actually. Once you have your core product to scale, it’s time to introduce API’s in your product. For you at Eventbrite, when did people start using API’s and when did you start exposing some of the API’s?
Renaud Visage: The first use case came from our users. They were spending a lot of time taking the data that they entered on our platform and copying it into event discovery sites, there were a few at the time UDO even for Yahoo, I think, different products. We built an even search API that allowed our partners to pull data from us basically in a very automated fashion. We had a daily chron that would pull the data from us and publish the events on their platform as well. It was a distribution API in a very rudimentary form, and it allowed us to be present in other properties than our website.
Renaud Visage: All of a sudden, all our events were automatically published to multiple places, which was great to acquire other customers onto the platform to buy tickets. Then we had clients asking for better reporting capabilities, so instead of building everything ourselves with the limited team that we had, we just opened up the API to access attendee information. It was private to the organizer, and they needed to have the right credentials for it. But again, it was driven by the needs of our customers.
Romain Huet: Sure.
Renaud Visage: I think now, it’s pretty much widely not adopted as a practice to build your core product onto API’s, but at the time, it was definitely not the case. We built an API that was in a separate file, that had no relationship with the product except the data underlying the storage of the client information.
Romain Huet: So you had to re-architect a little bit the way Eventbrite worked to have an expose?
Renaud Visage: We had to thoroughly re-architect over time, but for the longest time, all our API’s were in one file.
Romain Huet: Wow, that’s awesome.
Renaud Visage: That’s pretty scary when you unbundle that file and create the next version.
Romain Huet: MVP, it has to work.
Romain Huet: The original version of Stripe was just this one API to really accept Gull payments online. I think one of the first breakthrough that Stripe made was this idea of, Stripe.js, which was this idea to do PCI compliance in the browser, because if you rewind just 10 years, payments online before Stripe were extremely complex because as a merchant, if you wanted to sell any item online, or even just a service, you had to apply for this merchant account. It would take days or weeks to even get approved, but you also had to go through all these hoops of PCI compliance.
Romain Huet: For Stripe, we had normally this idea of, this has to be much simpler, you can sign up in minutes, but you can also start accepting payments without going through all these craziness of PCI compliance processes. So first, I would say the API and Stripe.js for what we call tokenization of call details in the browser.
Renaud Visage: I think at the beginning of your platform strategy, you’ll find that your clients have needs and that you can fulfill them with the right integrations, the right partners. So engage with them, understand how they run their businesses every day to see what first integrations make sense and what API’s you would have to build. You don’t have to make them public or open. We decided to make it open because we wanted everyone to have access to the content that was generated on the platform. But I think every business will have different use cases and just understand that as soon as you make an API public, it’s a long-term engagement, because you’re going to have people building other businesses on top it and they’re going to rely on access.
Renaud Visage: I think there’s been many examples of companies being built on top of just one partner, Facebook or Twitter, for example, then they were denied access for whatever reason, and their business failed. So understand the dynamics between you and your partner very early on so that you can have enough diversity in your business model to not be fully reliant on one partner.
Romain Huet: Yeah, totally. What would you say was the most surprising use cases for Eventbrite at the beginning when you started to open up these API? What are the things that you saw and you did not expect?
Renaud Visage: Well, I think we always underestimate the creativity of developers and what they can do with your data. We had so many hackathons happening and using our data for wide variety of uses. A lot of businesses started being created on top of our data, to fulfill some of the things we were not providing to our clients. Slido, is a good example of a company that started because they could connect with our API and gain access to all the attendees and engage with an audience, and I have countless examples of these type of businesses that today rely very much on us and our API’s to function and provide the service that they provide.
Renaud Visage: That was the very fun part of seeing what people can do without data. That I think was the most surprising. We thought we had the whole industry nailed down and fully understood, but then all these products started to emerge and filling very specific needs for specific clients.
Romain Huet: Yeah, what I always like to see when you open up API is like one of the creativity of developers and what they built with those and not only that, but also the feature requests they tend to have. Like back in a few years ago at Stripe before we had this product called Stripe Connect, for instance, which powers marketplaces, we had some very interesting feature requests. Some developers would ask us, “hey, I’d love to have thousands of bank accounts inside my one Stripe account” and we’re like, but what are you really trying to do?
Romain Huet: Then we realized that what they were trying to accomplish was actually marketplaces, they wanted to not only be able to accept money, but dispatch these payments across multiple sellers on their marketplace. That led us to build Stripe Connect to open up the door to all these marketplaces that we’re going to have to build before that. So always interesting to see feature requests from developers and how they push you in a certain way.
Renaud Visage: Also, you get lots of ideas from your customers, because our first users for the APIs were our customers. I remember going to General Assembly, which gives classes to a lot of people on technical matters, and seeing they had hundreds of classes and they built this very sophisticated reporting system on top of our API’s that they use every day to manage their business, and it was quite impressive and gave us a lot of ideas of what type of reporting we should be adding to our core product.
Romain Huet: Yeah, and sometimes what’s interesting is if your product is so massively successful sometimes even without opening official API’s, developers find their way to use your platform. I think I always like to remember that in the early days of Lego Mindstorms, the Lego company never had the intent to open up that to developers to play around, and same for the iPhone. Like Apple was not a platform company at the beginning. They thought of the iPod and then the iPhone as a finished product and they didn’t really have SDK’s and platforms until a bit later with the App Store.
Renaud Visage: That’s why it’s important to keep the pulse with your users in general and understand what they are solving with your core product. What you’re not solving for them and how you could be solving it through just integrations, because the reality is organizing an event has 2000 categories of businesses that are involved in one way or another and we couldn’t tackle them all, even if we wanted to. Even now that we’re a public company, with more than 1000 employees, there’s just too many verticals inside that daily life of the event organizer to do them all. We have to be very strategic about what we build, what we don’t build, where we partner, and what makes the most sense, in terms of ease of use of our overall product.
Romain Huet: I think that’s a good segue to our third point. We just mentioned how you start introducing API’s on top of your core product functionality. Now I think the next step when you start having some developers and some foreigners building, I think the next step is really to think about what kind of developer experiences do you want to offer to facilitate the onboarding and also, how to successfully start adopting some platform thinking at the company because now you’re going to have this API in the wild and some developers and partners are starting to build with it. So how do you start building a great developer experience I think is the next step.
Renaud Visage: Well, for us it took a very long time, I have to admit much longer than we should have. We launched a bunch of API’s, we had pretty bad documentation for them, people still built on top of it, but we didn’t really have a platform strategy at the time. I think we really started thinking about our product as a platform two three years ago. We started in 2006, so it shows you that you can go a long way without having a great developer experience, except that expectations have changed a lot.
Renaud Visage: What was okay in 2008 is not okay anymore today, because you have great platforms like Stripe, that have set the bar very high as to what kind of experience you should have when you’re a developer trying to integrate, and you have to at least offer something that’s in the vein of what’s available out there. About two years ago, we built a team around that. We redid the way we produce documentation so that it’s machine readable. We had to rewrite all API’s to make them more consistent. We entered to introduce concept of versioning, communication between us and the developers was very ad hoc before, not centralized, not shared.
Renaud Visage: There was no discussion group for developers to share solutions. We were not providing solutions, we were providing a list of endpoints and up to you to figure out what to do with them. So all these components, I think, eventually make the success of the platform. Once you have a true developer centric team that’s actually talking to customers every day and customer being the developer, in this case, understanding their pain points, what’s easy to use, what languages they use. We were talking about that. How do you provide the right tooling for them to get started as quickly as possible? How do you explain what’s possible with your endpoints and such as how many endpoints today? You don’t even know, right?
Romain Huet: I think we’re in the hundreds now with endpoints and API’s.
Renaud Visage: Well, we have hundreds as well. So how do you make sense of it for someone who doesn’t know anything about your business, because a lot of integrations are not done with specialists in your field. Like we integrate with MailChimp, for example, they have hundreds of different use cases, they don’t know much about events. So how do you explain all this terminology that you take for granted? How do you make it ingestible very quickly, and how do you make it as efficient as possible for a developer to get started playing with it? I think that’s the key for successful integrations.
Romain Huet: Yeah, I think we like to think about developer experience at Stripe as a constellation of details that really all matter. I think for us, and for me, the key is really to think about the friction points, like from the moment someone discovers Stripe outcome for the very first time, all the way to having a very successful business running life payments, like what are all the touch points during that journey, and where can we remove friction as much as possible. I think a lot of those things matter. There’s you mentioned documentation, you mentioned onboarding, but having simple apps, having client libraries and all the languages that you use, and frameworks also that developers tend to embrace these days. All of these things really matter for developers to be efficient.
Renaud Visage: Yeah, especially as you make the transition from having a few API’s here and there to really thinking about your company as a platform company. I think that’s the biggest transition in terms of organizational structure, and having enough resources dedicated to the platform side of things. But that also requires changing how developers develop, because documentation is not an option anymore. It’s a requirement. API definition has to be consistent with what’s been produced in the past. You have a lot more, I think, things that slow down development dramatically when you go into a platform strategy. Being conscious about that additional costs you will incur as an organization to get there. It’s non-trivial. I think we sometimes underestimate a lot what it takes to be and support and maintain a platform strategy.
Romain Huet: Yeah, and for a company like Stripe, I think the way we see what we’re building is really about economic infrastructure. So for all of you in the room building SaaS companies, what we really want you to focus on is really what you’re best at. What are the unique product features you can offer to your customers? You should spend as little time as possible on anything payments, that should be something that you delegate to some company like Stripe, for instance. We tend to think about our API’s like infrastructure.
Romain Huet: Infrastructure is they tend to be kind of decade undertakings. That’s why we want to make sure that if you start building on Stripe today, that should still work in three four or five years from now without you having to change your integration every week. It’s something that has to be robust for you to trust the platform and the infrastructure you’re using.
Renaud Visage: It’s also building all the components of the infrastructure. Endpoints are the start, but you need the reporting on performance, on number of calls you’re making, on how many you have left, on how much you’re going to be builds, reliability, webhooks, something that’s quite complicated to build, and really important if you want to avoid excessive usage over your API’s. These are maybe the more invisible pieces of the infrastructure that actually take a lot of effort to get right.
Romain Huet: Yeah, especially if you want to have some real time components in your platform. Maybe we can mention trust also, because as you’re having these API and you’re putting API’s out there when people start building with them, assuming they have a good experience with your documentation, your client libraries, you also want to make sure you build that trust over time. How have you been thinking about the trust with developers at Eventbrite?
Renaud Visage: There’s I think different levels of trust. We have to trust the partner that they’re going to deliver quality experience for our users, because once you have an integration, you’re extending your product beyond your core with something that you don’t fully control. So having enough monitoring of the performance, of the engagement, of how people are using it. Is it effective? Is it a good workflow? Being very critical about that so that your customers don’t end up distrusting you because you provided a half assed solution for them. Fully relying that the API will be available, that it won’t change without notice, that it’s going to be performance, that webhooks will arrive, all these things are expectations that if you do it right then you build trust over time, because they know they can rely on you not changing your policies all the time.
Renaud Visage: I think that we have many examples of large companies having to change access to certain API’s and kill a lot of businesses in the process and get a lot of anger from developers. So think pretty carefully before you open something up, because if you have to take it out at some point, it’s a big hit on the trust that the developer community gives you. Then, be fair, transparent and communicate a lot. I think we under communicate in the early days, because we didn’t have many people dedicated to the effort but you have hundreds of developers who want to talk to you every day. Make sure you have the capabilities to handle that.
Romain Huet: Yeah, absolutely. I used to work at Twitter before Stripe and I think communication was also something we could have improved. We became better over time, but back in the days of Twitter clients, for instance, when they became kind of a threat to the company, and like how the ecosystem was evolving around them. I think communication is key to make sure developers are building the things that you want them to build.
Renaud Visage: You have some big API changes thanks to regulation as well.
Romain Huet: Yeah, at Stripe. We mentioned versioning earlier, so we try to be always backwards compatible. When you build on Stripe, that integration will work over time, but what’s interesting is the trends of payments. I’m sure most of you in the room have heard of changes in regulation coming in Europe in September called SCA, like a strong customer authentication. Large fraction of payments in Europe are actually going to be triggering 3D Secure, for instance, starting September 14.
Romain Huet: So how do we build an API that is still extremely simple to use, but also making sure that all businesses have as little work as possible to benefit from that? We’ve been working on these new API’s at Stripe to make sure that whether you’re building kind of a SaaS business or if you’re building a marketplace, making sure that all this like, 3D Secure complexity and when to trigger it is taken care of by the platform itself.
Renaud Visage: But it’s a big effort.
Romain Huet: It is a big effort because that’s one of the rare cases when we communicate proactively with businesses to upgrade their integration to make sure the payments still work come September.
Renaud Visage: Otherwise, payments gets declined, right?
Romain Huet: That’s totally sure banks have to decline payments if they’re not 3D Secure enabled. It’s pretty scary but luckily, we’re on top of things. In terms of the developer metrics or actions, is there any particular things you look at or track in terms of the success of the Eventbrite API?
Renaud Visage: There’s the success of the developer trying to integrate and for that we look at time to first integration. We have a goal of making it as efficient as possible to start playing with the API because we think developers don’t like to read documentation really. They just want to download something and see what it can do so we try to focus on that. On the business value of the platform effort, I think we look at our platform effort as a retention tool. We think that once a customer has connected with different other products beyond ourselves, that they will stay around longer, and that it will be much more difficult for them to leave.
Renaud Visage: We look at how many … The percentage of customers that have more than three integrations through our hundred plus selected partners that we work with. That’s the one key metric we use. We don’t use monetization as a key metric because we don’t think it’s as relevant yet. Eventually, we’ll probably have some monetization efforts around the platform. But currently, it’s a way to keep people on the platform for longer.
Romain Huet: Great. Yeah, for us, we tend to obsess about simplicity of the API, especially as many businesses, in particular SaaS businesses are becoming a lot more sophisticated than they used to. Like five, six years ago, a SaaS business may just be, “oh, I want to charge like $19 a month and that’s it”. Now the business models have become quite advanced. People will want to have like, metered billing, tier pricing, still some usage records on Stripe, so as a result of that, we’ve made our billing stack a lot more advanced without obsessing about how do we make sure the initial onboarding remains very simple. I think that’s something we look at. The time to first payment and time to first charge on Stripe.
Renaud Visage: Organizing complexity to keep it simple is not easy.
Romain Huet: Exactly, doing simple things is actually pretty hard. Maybe we should move on to our fourth step, which is now you have this like API in the wild. We’ve covered how to think about developer experience and how to think about platform thinking. Now the first step, you start having an ecosystem. How do you build a healthy ecosystem around your platform and start leveraging your network effects? Maybe you can tell us a little bit about how you’ve thought about that for Eventbrite and, which side of the platform you started with?
Renaud Visage: We started with supply of ticket inventory first, that required customers and as I said, I think extensions of your products is probably the first area of the platform you build up, because there’s already an ecosystem of products that your clients are using. They want all of them to talk to each other, so that makes sense as the first natural extension. Today, I think we have 100 plus strategic partners, companies like Salesforce, MailChimp, Microsoft, and a few others, that are key software that our clients use every day that we need connected.
Renaud Visage: Then for us it was, how do you build the flywheel and how do you bring more consumers to our event organizers because in the end, their goal is to fill this room with as many people as possible. For that, we’ve really invested in finding the right distribution partners. At your time it was event flow and the event discovery platforms. Today, it’s much more specialized audiences like Spotify, we have integration with them. Or we want to be where the consumers are so we have a lot of integrations with Facebook, YouTube. If you find an artist that you’re interested in, you’ll find the events that we power that are where this performer is going to be.
Renaud Visage: Distribution partners, we have about 50 Plus I think today. Then there’s the products that are purely built on top of our platform that don’t access our users. That’s a new area of investment. The first one we launched was Facebook ticketing. If you go on Facebook in the US, today, you can start selling tickets directly when you create your event on Facebook, but we power all the backend for that. So that’s a new area for us to invest in, really that has the most potential, I think, for us to grow in the future. Especially with partners like Facebook that have millions of consumers and millions of events created every month, like powering their pay tickets solution was a big deal for us and really gets us excited about the future of our platform as a company.
Romain Huet: That’s awesome. I think it’s the same for Stripe because even though we started with the utility API to make it super simple to accept payments in any apps and services, the ecosystem sites are going to self organize around us. Like when Stripe grew in reach, I think over time a lot of developers started to build, not only use Stripe for their own apps, but also a way to extend what Stripe can do for themselves. So we’ve seen like thousands of companies and as you pointed out, I think there’s different kinds of banners in that ecosystem. I think we’ve seen the ones where, companies like Shopify, DocuSign, Zero, they want to work with Stripe to build an entire platform and let businesses accept payments.
Romain Huet: But we’ve also seen integrations, extensions or plugins, if you will. Extensions, for instance, like segment or … Their goal is not really to accept payments per se, but rather extend the value of Stripe by having these data interfaces, like how can you connect your stripe account to a CRM for instance, all of those things. It’s interesting to see that ecosystems tend to first, at least for us, like self organize, and then we had to organize this whole ecosystem, when we reach kind of thousands of companies integrating with us.
Renaud Visage: Yeah, and then you started building products on top of that.
Romain Huet: Exactly. Yeah, absolutely.
Renaud Visage: So you’re on a reverse journey compared to us. We started with the product, we’re evolving into a platform. You’re a platform from the beginning evolving into multiple products.
Romain Huet: Into multiple products including like, radar for machine learning and FAR, then all of these things. One thing we also did just last year was to launch like a pilot program because for the longest time, we had all these partners that we had identified and tried to push our users toward if they wanted to have like specific needs. But with the Pilot Program, our goal was amplify and make sure we grow the reach of all of these companies too that do the effort of integrating with the Stripe data.
Renaud Visage: It’s if its like a acquisition business. You have the organic growing side of the business where people start building on top of you, and then on the other side, you need to think about your strategy for the platform. Like which are the most valued partners that are currently not on the platform that you need to onboard? How do you make the current integrations better, more efficient? How do you monetize them down the line? All that goes into when you are more mature platform in the end.
Renaud Visage: You have to have a strategy around, what’s more important to build because you still have limited resources so you cannot go after every partner the same way. So being very strategic, having the right people on the business side. Partnerships are a two way conversation, you need to find the relevant ones and find the right economic model to keep them engaged over the long-term. All this requires thinking of it as a very different organization in the end than your product organization.
Romain Huet: Always be transparent with those partners. If it comes to time where your product will start doing natively some of the features that they had previous to build. I think it’s really important for the trust with the partners to communicate proactively ahead of time with them. I think that brings us to our last step, like when you have this platform and this flywheel, as you mentioned, really running nicely, you already have to start trading your platform like a core product, because at that point, it’s become maybe your most important product, in fact. So how do you think of that at Eventbrite? Do you think you’re at that stage or you’re still transitioning to that stage?
Renaud Visage: I think we have all the pieces yet, but we do treat it like a product. We have the same rigor in talking to our customers. In this case, our partners, our developers, all the different actors of the ecosystem, engaging with them regularly monitoring all the KPIs around the platform side of things, trying to think creatively about what solutions we can build on top of the API’s we have or the new API’s we’ll build. Then, seeing what the gaps in the market are, and do we solve them through product or through platform?
Renaud Visage: I think that’s probably the biggest decision that’s quite hard sometimes. Do we build or do we integrate? I think we’re almost there. I think we need a few key hires to really think about monetization because we haven’t really talked about that that much. We have some deals with some partners, but it’s not something we do consistently. Then imagining what the future revenue lines could be, that would be fully powered by the platform like the Facebook ticketing deal. How do we have more of those?
Romain Huet: What kind of mistakes do you see companies make usually when they approach their platform, because we’ve seen even the largest companies like Google can sometimes fail at a platform play? Like Google Plus was one of them. Unless you’re 10x better with your network effect, you can have a platform that fails. What mostly top of mind for you in terms of those mistakes to avoid?
Renaud Visage: I think maybe something relevant for this audience if there’s a lot of early stage startup founders here, it’s doing it too early. I think once you’re … If you don’t have the critical mass of an audience, then doing the platform will distract you and will not result in the type of benefits that you would get from doing it later once you have that critical audience you can address through a platform offering. Treated it … Not eating your own dog food, that’s probably the biggest mistake we made. When we launched our API, as I said, there were separate file not used by any of our product, and inevitably when you do that, it becomes dated on the first day that you launched basically.
Renaud Visage: Any new feature will not make it into the API down the line. So we ended up with functionality that was totally not available through the API. Still the case for some API’s actually, but we now changed completely the way we build software. It’s all API based internally. We expose some to partners, we expose other endpoints to the general public but we have a strategy around it and everything we build is openable at some point.
Romain Huet: Yeah, I think being API at first is a real change in culture, too. If you’ve not had this culture from day one, thinking about, no, I’m going to build the API’s first then I’m going to consume my own API’s instead of doing two separate things. That’s the only one.
Renaud Visage: I think something we talked about, maybe the next evolution would be to build your front end the same way. Think about how you build software, you building widgets basically that you put on your own sites that allow you to do a certain thing like creating an event or buying a ticket. What if that thing was a widget that anybody can put on their site to do exactly the same thing? I think you did that with Stripe with the first card payment widget that you launched?
Romain Huet: Yeah, and same like relevant for SaaS businesses in this room. I think we’ve seen time and against so many SaaS companies having to reinvent the wheel for so many things. Like, I want an invoice and of course, I want that invoice to be nicely designed and customized and work well with every single email clients, and I want my customers to pay for these invoices with cards or offset by debit and all of these things. There’s certainly so much time spent for SaaS companies that we actually took on all these efforts to and to provide a lot more UI components so that people in the room here don’t have to reinvent the wheel, just like customize a little bit what makes sense for them.
Renaud Visage: I know we’ve done the power of widgets for a long time. We’ve had widgets pretty much since day one and they’re very powerful because they abstract a lot of the complexity of the underlying API’s and give you full control of the experience so that you can optimize it much faster and with more power than any of your clients. I think that’s an area where we’ll see a lot of investments from the smart companies providing much easier to integrate tidbits of code that you could just put in and get started.
Romain Huet: I think it’s important to have the full range too, like we go from UI building blocks at Stripe that you can customize to every pixel and then we have like those invoices and Stripe Checkout products, you can just like drop in and customize, just like colors and logos, kind of have this full spectrum for people to customize. In terms of trends, by the way in terms of platforms like we mentioned, for instance the Stripe side there’s trend around authentication for payments and everything. Any particular trends you’ve seen or things you’re excited about for the future of platforms?
Renaud Visage: I think there’s a lot of place for, like decentralized services that you can integrate into your platform. Products like fraud and risk. They should be a few we can just plug and play and not have to reinvent the wheel. I think we spend way too much time rewriting things that others have written many times. I’m curious what’s going to come up in our ecosystem to solve these problems.
Romain Huet: That’s always been what’s top of mind for Stripe. Like every time we see a company or multiple companies, in fact, rebuild things we’re like, no, we should definitely do that for them and take the burden on us so they don’t have to spend their time and energy on these things. We’re almost at time but I guess maybe we could do a quick recap of what we covered. So if we were to recap the five key steps of turning your product into a platform. We said first, prepare your core product first for scale, especially if you’re not a platform first company, but if your product first company.
Romain Huet: Second, introduce the API’s and start exposing some of your core features on this API. Third step, think about developer experiences and make sure you remove all friction possible for your developers and partners to really have a great time being successful and adopting your platform really quickly. Fourth step, we mentioned how to create a healthy ecosystem around you and leverage your network effect and finally, really treating your platform like a core product because it might become ultimately the most important product you have and that requires nurturing every day. I know there’s a lot of SaaS companies here in the room, any parting thoughts for you, Renaud, on what you would advise them if they want to aspire to become a platform?
Renaud Visage: I think it’s just very difficult to iterate on the platform side. Think about it for the long-term, engage for the right reasons, build the right team around your platform side of things. It’s not like a product where you can launch new features every week, so be thoughtful of that. Think of the long-term strategy and how it reinforces your mode for your customers, and then see what are the relevant partners for your offering because every solution is going to be different.
Romain Huet: Cool. Well, thank you, Renaud and thank you all of you for joining us today. We’re just at time now but we’ll be around for questions. Thank you.
Renaud Visage: Thank you.