Learn how Jed Ng and his team built the world’s largest API marketplace, scaling Rakuten RapidAPI to over 8,000 APIs and 500,000 developers.
Davis Baer: What’s your background, and what are you working on?
My name is Jed Ng. I work on building and scaling new tech ventures. I’m especially attracted to the ‘0–1’ of creating new market segments. I also work by bridging corporate/startup partnerships.
I’m currently the Head of Rakuten RapidAPI- a venture born from a deal that brought A16Z-backed RapidAPI and $10BN+ eCommerce giant Rakuten together.
Rakuten RapidAPI is an API Marketplace. Our mission is to empower software developers to build game-changing applications through web APIs. We’re building a single platform to unify the fragmented Developer experience (DX) when it comes to web APIs. This means connecting their entire user journey- from API discovery to testing, service subscription and payment, management and analytics.
We’re the world’s largest API Marketplace with over 8000 APIs and 500k developers. Some of the biggest names in the API economy are already on the platform including Microsoft, Sendgrid and Crunchbase.
What motivated you to get started with your company?
I previously helped establish Ericsson’s cybersecurity practice where our core team booked $15M revenue in our first operating year. This experience taught me how large corporates bring leverage in accelerating new ventures (if correctly set up). I wanted to apply this lever in a startup/corporate model and looked into APIs as an interest area for a potential executive sponsor.
My thesis of the API Economy was:
- APIs are useful: Application Programming Interfaces (APIs) are incredibly useful things for software development. Their ubiquity in the digital economy (mobile apps, cloud services) proves that. The industry had matured to the point that architecture largely coalesced around REST standard.
- Digital fragmentation was a big source of market friction. The ‘state of the art’ in the API economy today is ProgrammableWeb which solves API discovery but that’s as far as it goes. Large, established tech incumbents command asymmetric API traffic volume, making it hard for new APIs to be successful. New, up and coming API services were lost in a ‘digital ocean’
- Persistent market gap: API solutions and IT spend was pouring into lifecycle management and gateway solutions like Apigee, CA Technologies, KONG and IBM Connect. There’s even a Gartner Magic Quadrant covering this segment. These solve the problem of creating and managing internal APIs, but do not help developers discover, connect to and manage public APIs.
- Validation: Companies like Twilio, Stripe, Nexmo had spent over a decade proving out API-based business models and had validated APIs as a distribution technology and monetisation model. They were hitting unicorn valuations many times over. They were THE hottest tech brands. This was proliferating a wave of API-based businesses that exacerbating the digital fragmentation problem.
API Traffic Distribution: Pie chart based on Apigee’s State of APIs 2016 report
The market gap seemed to revolve around ‘making the market’ between API Providers and Developers via a platform. Such a platform would abstract many of the administrative painpoints such as key management, multiple bill payments and dashboards; simply put, we needed to build the “App store for APIs”
- Product: Developers needed to be able to able to discover APIs for specific services/functions, test them, purchase and pay for subscriptions, and manage them from a single interface.
- API selection: Developers need to be able to access a broad selection of web APIs in order to be able to find something for their specific use case or vertical, rather than to have to research and deal with specific API providers separately.
There are a few characteristics of this business worth highlighting:
- Multiple scaling levers: As developers spent on platform, they were more likely to find needs for new APIs. As new APIs are added, we’d be able to offer them to existing users. As their applications grew, so would connected API traffic and we’d grow with them.
- Platform stickiness: Absent a platform layer, dealing with multiple APIs from different sources creates administrative pain for developers. They’d have multiple billing accounts, dashboards and interfaces to manage on an ongoing basis. A platform layer solves these pains and creates user stickiness as the number of API connections grew.
- Same supply and demand players: Both API Providers and API consumers are Developers. This is immensely powerful to building network effects, not to mention a unique one. I can think of no other marketplace where the same players exist on both sides of the platform.
The crux in multi-sided businesses lies in cracking the network effect problem- typically by subsidizing one side of the business. We solved this by forging an agreement with RapidAPI. This gave us an MVP product, 8000 APIs and 500k Developer users on day 1.
Rakuten RapidAPI’s mandate is to bring the API Economy to APAC where 37% of the world’s developer population resides. We’d have home turf advantages of brand awareness, footprint, and operational know how. Rakuten benefitted from a vastly accelerated time-to-market, low cost of entry, and de-risked opportunity. The aspiration is to scale up together in a bid to hit critical mass.
This entire process of ideation, navigating a multi-billion corporate, securing endorsement and funding, brokering an agreement and product development took 18 months.
KEY LEARNINGS: Corporate/startup partnerships can have extremely long cycles, require immense effort with uncertain payoffs. Play for the >$100MN opportunity; anything less is unlikely to be worth it. The value of a dominant, global API marketplace is at least unicorn scale. Check out this article by 500 Startups where I talk about this.
What went into building the initial product?
Our MVP was a localized/forked version of RapidAPI’s tech stack with tooling in place for developers to find, test, connect and manage APIs. Our major scope decisions revolved around localization for Japanese users. This involved translation, QA, pricing and currency, and user flow which we completed in 13 weeks.
Localization means much more than language. Understanding local user expectations is key. Japan has much higher expectations around quality, even for startups. In addition, “made in Japan” conveys value as well. We reflected this in ways like Yen-based pricing estimates, and adding the Rakuten brand name. In addition, we applied a higher translation standard to API documentation, particularly for top APIs and relied on machine translation for the long tail. The trick is in finding the balance between quality, cost and speed.
How have you attracted users and grown your company?
Since launch, we’ve focused on growing and scaling inbound traffic which would naturally convert into registrations and API subscriptions. Content generation backed by good keyword research is absolutely crucial.
We’ve used a few tricks to maximize impact in our content efforts. We created a content generation machine by defining limited content formats with standard repeatable frameworks that could be executed by freelancers. As we expanded, we created self-organized writing teams that could supervise themselves. We also employed link retargeting within content that created opportunities for ad impressions that drive ~1% CTR.
For offline, 3rd-party hackathons have been a great format for getting Developers to interact with the platform. We approach hackathon organizers to provide Rakuten RapidAPI as a perk. This reduces event preparation time, lets us interact with new users each time, and allows us to stack up more events on our calendar. We end up finding a few super fans at each of these events which we’ve spun up into a champion program that covers 3 countries.
PRO TIP: Always ask for pre-event webinar and presentations, and speaking slots at kickoff for hackathons. Without them, onboarding Devs who don’t know who you are and who are busy trying to get down to work is nearly impossible. This massively impacts your capture rate.
In our first 5 months, we had 30,000 unique site visitors from 110 different countries with a weekly average of 1500. We’re growing traffic at 20% month-over-month and are now turning our attention towards managing the lower parts of the funnel.
We also launched Rakuten RapidAPI 2.0 in December 2018. This was a ground-up redesign of our API Marketplace based on the collective experience of us and RapidAPI.
Rakuten RapidAPI 2.0 new UI design
Besides a cleaner new UI, we introduced a few features:
- API Collections: Groupings of functionally similar or complementary APIs. This structure functions almost as featured placement slots and let us target long-tail APIs.
- Performance Metrics: Each API now shows detailed performance statistics including Popularity Score, Latency, and Call Success Rates. This is incredibly meaningful in my view. Since Rakuten RapidAPI is a marketplace, we can be an objective ‘source of truth’ for Developers in deciding which APIs meet their requirements.
- Better information discovery: We’ve introduced a popover feature in our main discover interface that provides additional information without navigation. API Pages now also show sample code snippets so that Developers can see what they’re getting without having to configure an API call.
API Performance metrics
The new UI has resulted in >20% bounce rate reduction and the long-tail collections page structure is generating 5–10% additional site traffic.
On the other hand, it’s important to share how we could have done better. We spent money to be at general tech events as exhibitors/sponsors. The results range from low returns to abysmal failures. If you’re marketing to Devs, you need to be at developer specific events.
Growth hacking was something we should have started earlier as well. For instance, trying out targeted cold emails or webscraping potential users using Facebook groups.
Hiring and recruiting has definitely taken longer than expected. The candidate profiles we were seeking- bilingual, risk appetite for startups (a less respected career choice in Japan)- are especially hard to come by. We should have been more aggressive at the start.
What’s your business model, and how have you grown your revenue?
Our business model is completely based on a platform fee on top of transactions. This comes out of API Providers.
The site is completely free for Developers and we offer API services at the same pricing as by going direct to source. They receive access to a massive 8000+ API inventory (and growing), plus all the tooling such as testing, code snippets in multiple languages, single billing and analytics.
For API Providers, Rakuten RapidAPI is an additional channel to reach our 500k developer base on day 1. There are no fixed fees for maintenance or listings. We only monetise when developers subscribe to an API plan via our platform. In this way, we align our interests with API Providers and help them monetise on a long-tail audience which they would not typically have access to.
What are your goals for the future?
Our goals are simply to grow our developer base as quickly as possible. Our combined user footprint with RapidAPI of 500k represents around 2% of the world’s professional developers. There’s a long way to go. We’re also considering how and where to pursue market expansion by building nucleus teams on the ground.
We’re also planning to roll out an Hub solution for Enterprises in 2019. The hub functions as an internal API Marketplace that sits on top on existing catalog or gateway solutions so that enterprise developers can consume both internal and public APIs. Public APIs from our external marketplace could be whitelisted for internal consumption. The enterprise would also be able to seamlessly publish their own APIs for would external users for monetisation. We want to get initial customers this year.
What are the biggest challenges you’ve faced and obstacles you’ve overcome? If you had to start over, what would you do differently?
One of our biggest (and ongoing) tensions is on how much to focus on Japan vs the rest of the world. On the one hand, our team is based in Tokyo and we can engage users more intimately. On the other, there is a far larger English-speaking developer market around the world. Reaching them relies on digital channels which are lower-touch.
There are no easy answers and this affects team dimensioning and recruiting. My view is that focusing on Japan only solves Japan and learning to grow and scale digitally is a separate challenge. If success means getting to dominant scale as quickly as possible, that suggests taking the digital route. From a team point of view, building a team of talented professionals with the right mix of language skills has taken longer than expected. In hindsight, we should have started sooner and have been more aggressive with recruiting.
Another thing we grapple with is how big the gulf between corporates and startups truly is. It’s a big reason why these sorts of partnerships often fail. This applies in product development and QA standards, time-horizons, agility, user acquisition vs monetisation. We are constantly bridging these differences.
PRO TIP: Managing corporate and startup partnerships is all about parties having an equal voice. Having executive sponsors on both sides at equal footing is a great way to surface difficult issues for constructive discussion.
Have you found anything particularly helpful or advantageous?
In situations of uncertainly, the natural human reaction is to stop and assess. Being able to compel yourself to forge ahead is key to creating opportunity out of chaos. I apply something I call “the 70% rule” where I look for a 70% correct solution to a particular problem and move to execution. It’s a good heuristic when things are unclear and chaotic.
Look for opportunities that have natural tailwinds or rising tides that will push you forward. In my case, there were 2 tailwinds. Firstly, there was the explosive growth of API-based businesses like Twilio and Stripe as catalysts of new API businesses. Second, there’s a rise in software developers as a user segment. This is manifest in GitHub acquisition by Microsoft and GitLab $1.1 BN funding round, both of with happened within 4 months of each other in 2018.
What’s your advice for entrepreneurs who are just starting out?
Entrepreneurship is all about execution. Surround yourself with people who are ‘do-ers’. They’re a great way to stay motivated to keep moving forward. Also, don’t be afraid to ask for help from mentors and advisors. People can always say no but you won’t know until you ask. Good mentors- the kind you want to surround yourself with- tend to say ‘yes’ if you approach them correctly.
Great resources:
- Fundraising: Kauffman Fellows Venture Deals course (FREE)
- Growth hacking: https://neilpatel.com and https://www.producthunt.com/posts/the-bamf-bible-2019
- Programming skills: Edx.com
- Getting found by early tech adopters: producthunt.com
Where can we go to learn more?
Rakuten RapidAPI is the world’s largest API marketplace with 8,000+ web APIs and used by over 500,000 developers. We enable developers to build transformative apps through the power of APIs. Find, test and connect to all the APIs you need in one place!
Check out some of the world’s best APIs including Microsoft, Sendgrid, Crunchbase and Skyscanner.