Ryan Sarver is a Platform Guru with years of experience building and scaling platform businesses. He is currently a Partner with Redpoint Ventures. Prior to joining Redpoint, Ryan was Director of Platform at Twitter, where he oversaw the Twitter Platform and its ecosystem of companies. I recently had a chance to talk to Ryan about platforms and what he learned from his time at Twitter. A transcript of the interview is below.
AM: You’ve been at Redpoint Ventures now a little over six months, and we saw that you were at TechCrunch Disrupt recently. Were there any trends that you saw there?
RS: One of the benefits of a firm like Redpoint is you can see a really wide breadth of things. I think there’s some obvious major trends that people talk about often. You’re starting to hear a lot more about Bitcoin. Another trend is taking stuff that builds infrastructure of some kind, like Twilio or Stripe, and making it available on a broad scale to developers to lower the friction of integrating that feature into their platform. That’s one category that Redpoint has done really well on.
Another one that I think is a broader trend is marketplaces and the evolution of how marketplaces have changed. We’ve done things like HomeAway, Beepi and Homejoy.
There’s been a change in marketplaces from the Web 2.0 days – where companies like Craigslist and eBay, which were really just kind of an unfettered peer-to-peer marketplace trying to connect two sides – to today’s current marketplaces where they take on ownership of more of the transaction to guarantee a better result. So Homejoy is actually training and licensing all of the home cleaners. They’re going to guarantee a much better outcome and a happy cleaner instead of just going to Craigslist with a home cleaner that is not vetted and not rated, and where you have no idea what you’re getting into.
AM: One of the things that seems to be particularly difficult for these platforms which revolve around services provided by humans is user acquisition, both on the consumer and producer side, with the producers being the folks that are providing those services. How does that world shake out for you?
RS: Well, I think we all know the major shift is toward mobile. We have these devices in our pocket that are with us at all times and always on and connected. This is what enables a lot of these marketplaces to exist, really. I think you’re seeing that for a lot of them. Uber, Shyp and Postmates are good examples of marketplaces that couldn’t exist without our current mobile devices.
But even with mobile, you still need to build up two sides of the marketplace. You need to connect people wanting Uber, people who need a ride, with the people who can deliver that ride. It still comes back to the simple question of “How do you acquire those two user groups?” Mobile made the marketplace efficient for both sides and created a great user experience, but it doesn’t really help on the user acquisition side as much. So companies are still having to figure that out.
There’s a limited window of attention for you to draw in users and an even more limited set of users that are going to stay with the service. Maybe once you’ve heard of an application, mobile has made it easier for those users to find and connect with your app, but I still think that early user acquisition is one of the biggest things that every company we work with and talk to struggles over. There are a million applications just on the App Store, and it can be incredibly difficult to raise above the noise to acquire users.
AM: You had experience going after that producer user base when you were at Twitter. As Director of Platform, you led the introduction of the developer platform on top of the Twitter social networking platform. How do you feel that the introduction of the development platform helped to support Twitter’s growth in user acquisition?
RS: In it’s early days, I’m not sure it changed Twitter user acquisition. It did a little bit, but I don’t think it was one of the major factors. I think what it changed for people was how you could engage with Twitter once you had joined, which then leads to stickiness and longer-term engagement. If you remember the early days of the Twitter platform, it was mostly about third-party clients. Twitter was just a website and SMS and it could barely stay up each day.
The development platform was something that allowed people to develop clients for mobile and for desktop and for Linux. They created different web versions and the myriad other ways that people could then connect to Twitter. It allowed Twitter to stay stickier in users’ lives because Twitter was now available wherever they were. So, if mobile was taking off and Twitter didn’t have the resources to go build a mobile client, other third-party developers built out these mobile clients. And if Twitter didn’t have some kind of mobile presence at the time, I think we would’ve lost a lot of those users in the early days.
I think the development platform shifts its value to the company and to the broader business depending upon the platform’s maturity level. In the early days, Twitter was building its own awareness through the media coverage and word of mouth. The third-party client really helped users stay on the platform and engage with Twitter in ways they wouldn’t have been able to if it was just Twitter sending their phone’s tweets.
AM: These platform business models are making their way into many “traditional industries.” Are there any dynamics or differences that you see when operating in the pure tech world like you did with Twitter versus a more traditional industry like you do with Homejoy?
RS: If you look at marketplaces today or developer platforms with Twitter, they’re really both two-sided businesses and there’s a lot of commonalities around how you manage those. The job at Twitter was to connect a user with a third-party developer to be able to give them some new functionality – either get them to join Twitter or get them to stick around longer than they otherwise would have. So, our job was to build up both sides of that market and connect those users really efficiently. Marketplaces are very similar where you’re connecting someone who needs home-cleaning services with someone who provides that service. And, you need both of those sides to connect really efficiently in order to build up both sides of that marketplace.
There’s a lot of lessons to be learned from both of those things and they have a lot of the same principles in how they operate. So it takes someone with my background in dealing with developer marketplaces that can quickly jump over to talking to the Homejoy people either during a pitch or a $2 million investment to help them think through some of the things that we learned in a somewhat similar but different environment. We used to talk to guys who ran the eBay marketplace. They were trying to get power sellers on the platform so that consumers had more options to buy. When you talk to them about how they did it and the rules they put in place, it was very similar to the way that we managed the Twitter ecosystem. It was really uncanny. So, I think there’s a lot of lessons to be learned that apply to both types of platforms.
AM: Can you talk a little more about how rules and standards helped shape that platform? How it helped give trust to either side of your marketplace and enabled the right kind of interactions between your two different sides? You mentioned it briefly and we think it’s a very important aspect of how these platforms really grow to a point of maturity.
RS: The way I used to talk about it to our team at Twitter is that we’re the mayor of a town. Our job is to create incentives and disincentives to produce the best behavior, the best outcome, from a bunch of people you’ll never meet. Remember, these aren’t small marketplaces. On Twitter, there were three or four million different websites we were working with and maybe a million different developers. So it’s huge ecosystems where you’re only going to meet a very small fraction of the people. So our job was to create policy that lets people know where the guardrails are and what behaviors are expected out of them. I went into the job having never done any real policy work before and I never realized how important policy was going to be and it ended up being a huge part of our time.
My advice here to anyone trying to grow a platform would be to think through what you want the desired outcome to be and try to find the most explicit way to write the policy to produce that behavior. If you become lazy or take broad strokes with your policy, you create a lot of unintended consequences, usually in a negative way where you disincentivize certain behaviors that could have been really positive for your platform. So at Twitter getting this right was something we spent a lot of time on. We made sure that we put that time and effort into crafting a good policy that only took out the bad behavior we were explicitly trying to strike, without taking out any of the positive side.
AM: Right. And those policy decisions have ripple effects across both sides of your ecosystem, so it’s important to get those things right on the first try.
On a similar note, what are some of those stickiness factors that you see for developers? How do you draw them to work on your platform?
RS: Well, that’s the biggest question right? I think a lot of it comes down to what the platform offers natively. Twitter was fortunate in some ways that the content we were offering, the tweets and the system, were this really unique data set that developers couldn’t get anywhere else. We talked about it as the “global consciousness” that was suddenly indexable and analyzable. It was just this amazing data set that you couldn’t find anywhere else. That alone started to draw people in because it was just something developers wanted to tap into and work with.
So just producing an API doesn’t mean that’s it’s interesting to developers. You have to start from their point of view and think, “What about this is going to draw people in and make them use this? Why this is unique in the world relative to other data sets?”
The two most obvious things that developers look for are money and users. So are we able to bring distribution? This is what Apple does incredibly well. If you are a developer in a garage in Palo Alto and you develop a mobile application, you can put it in the App Store and your app can just explode because the breadth and the distribution that Apple has is super unique. So that’s what I think developers generally look for, “How can someone deliver me money?” or “How can someone deliver me users?”
There’s obviously a lot of other secondary or maybe tertiary benefits as well. You know, “Are they promoting me well? Do I feel fulfilled? Are they recognizing my hard work?” There’s all sorts of other smaller things, but I think those things fade over time and the most enduring platforms are the ones that deliver fame, fortune and/or users.
In the end everyone wants to be able to work on stuff they’re passionate about, right? So you’ve got to figure out what’s the motivation for developers and then work backwards. I think a lot of people now just build platforms mindlessly and they don’t think through the motivation of their users. They think that a developer will just inherently want whatever the platform has because they’re making it accessible. Meanwhile the data set they’re offering might already exist a million times over, or there might be no real motivation for a developer to use that API or that platform.
RS: I’ve walked away incredibly impressed with Facebook’s moves. They make really bold, big decisions that are probably not obvious. Contrast that to Apple acquiring Beats, which seems like such a standard, uninteresting acquisition. A music business buying another music business that has a slightly different model. Maybe it’s because they have hardware we don’t know about, but on the other end of that, Google is buying DeepMind and crazy robotics companies. They’re trying to disrupt themselves. And I think that Facebook and Google are taking the biggest strides in doing that while Apple seems to be taking a protectionist approach to their existing business model. So I give Google and Facebook a lot of credit in their M&A strategy and Apple’s maybe to be determined. It might be too early, but it just seems like a very uninteresting acquisition.
AM: Thanks Ryan, and thank you very much for your time.
RS: Thanks Alex. Always great to talk platforms.
Want to read more? Check out our interview with platform blogger Sangeet Paul Choudary.
Filed under: Platform Innovation | Topics: Ecosystems, marketplaces, Platform Startup Advisory, platforms, Twitter
Platform InnovationRead more