John Musser on how to start an API developer community

John Musser is the founder of ProgrammableWeb, the leading online resource on open APIs. John is an industry expert on APIs, quoted in the Wall Street Journal, New York Times, Forbes, and Wired. He's a frequent speaker at conferences including SXSW, Dreamforce, and Web 2.0. He also consults with companies on API strategy and trends, with past clients including Google, Microsoft, and Salesforce.

API Economist:  You founded ProgrammableWeb back in 2005, is that correct?

John Musser: That’s correct. It’s coming up on its eighth anniversary this summer. It was really the birth of the open API and web mash movement. The phrase “web mashup” was really coined four or five months before we started ProgrammableWeb. That was the same spring when Housingmaps.com was built, which was essentially the first quintessential mashup (a mashup of Craigslist and Google Maps).

API Economist: Why did you start ProgrammableWeb?

John Musser: Out of need. I was researching a different business idea that involved the type of things that APIs seemed like they would facilitate. As I was doing my research I found bits and pieces of information on APIs, but never felt there would be a place to go. So the byproduct of my research, I registered the domain name, put up the site, and lo and behold, other people needed that information too. So it was really just born out of necessity.

API Economist: What’s changed since you gave your QCon presentation last March on the state of open APIs?

John Musser: More APIs [Laughter]! Certainly that's been true every year, but we've doubled, yet again, on a much higher volume. The numbers keep getting bigger, but the doubling keeps happening, in terms of the gross API market. So, it’s a pretty steep slope. This past year we've seen it in more and more places...more and more verticals...more and more sectors. And this includes the enterprise market. We've seen a lot more mobile use cases. There are a lot more people opening up and consuming API's as part of a mobile or device strategy.

APIs are a great enabler of cloud computing and all those cloud services mean that APIs will be the glue that stitches it all together.

API Economist: Looking forward over the next 24 months, are there any particular markets or sectors that excite you the most, and why?

John Musser: I think a lot of them revolve around the enterprise because I think that enterprise integration is unsexy. But as they say in the VC market these days, enterprise is still where the money is. The leaders in the enterprise have been opening up data via APIs, but we are going to see a lot more of the laggards opening up APIs as well as thinking more seriously about how to consume APIs. I haven’t seen much in manufacturing. I also think we need to take a second run at sectors like finance and telephony. Many of the Telco’s are starting to open up their back-end systems to developers. And then there are guys like Twilio who really did a fantastic job of leading the way in the communications sector.

API Economist: Regarding the API billionaires club, most of these organizations tend to be New Economy style companies. Do you see the enterprise breaking into this club?

John Musser: Eventually, yes. We started to see some of it again in the communications sector. These folks are starting to do pretty large volumes. That said, it’s going to take time. I think the enterprise is going to adopt it more slowly, but you are going to see those huge volumes from the get go.

API Economist: In the case of enterprises, do you think it's going to be more around the semi-private API or the nonpublic API for enterprises?

John Musser: I think that’s true. Look no further than Salesforce. They've said for years now that more than half of their traffic volume comes from the API, not from the web app. I think that demonstrates the amount of integration that has to happen on a successful cloud-based platform. This is the model that we will see a lot more of in the enterprise space. The SaaS (software-as-a-service) service is being integrated into the enterprise, and then the enterprise is thinking about uncovering what aid and services they should then in turn expose to their partners and customers.

API Economist: We’ve seen significant adoption of SOAP and XML in the enterprise. What about REST? Do you see it being adopted in the enterprise?

John Musser: REST is already being adopted in the enterprise. It's still new, but it'll become an implementation detail. I think this shows that in the broader market, REST has been a runaway success, at the expense of any other alternative, which has clearly not been true in the traditional enterprise space. Let me put it this way: if you talk to a lot of API providers, there isn’t necessarily a REST versus SOAP issue. Most API consumption tends to happen through a wrapper or an SDK around an API, so the majority of developers consuming REST-based APIs aren't necessarily exposed to the REST itself.

They're using a .NET, or a Java, or a Ruby wrapper around the API anyway. That insulates developers, and again a traditional stack like manner, from the underlining service. I think that will probably be the way that it happens in the enterprise as well.

API Economist: It seems that we are still in the early days regarding API adoption. What’s your observation?

John Musser: We're still in the super early days. Even if you look at the history of ProgrammableWeb and the web mashups from APIs, that's seven years, which, on the one hand we've come a long way, yet on the other hand, it's still a small fraction of the overall development pie. There's a huge chunk of the development marketplace that has not yet adopted these models either on the production side or on the conception side. Literally millions haven't, but they will.

API Economist: Why is it that some organizations and markets jump headlong into opening up APIs? It seems for some they are forced into it due to disruptive and hyper competitive market conditions. Retail is a good example of this.

John Musser: I think that's a really interesting question because one of the phenomena that we've seen over the years at ProgrammableWeb is how often these things fall like dominoes within a certain industry. If you look back over some of the predominant sectors of where open APIs or web based APIs are very common, you can see they've reached a tipping point or there might have been an inflection point where someone was very successful with an open API strategy.​

Google came along with Google Maps and disrupted that industry, for many reasons but one of them was the open API and the ability to integrate Google Maps anywhere. Now there's no such thing as an online mapping service that doesn't have an API. You could also look at video, that's essentially the same thing. There's no online video, photo, music, or entertainment service that don't have an API…they all do.

Social networks used to be closed things. They're still closed in many ways, but certainly without APIs. There was this website called MySpace. And then this other website called Facebook opened up an API. It made a big difference back in 2006 and 2007. You can graph the trajectories of those two companies going in opposite directions for many reasons, but certainly one of them had a (open API) platform strategy.

So we see again that you can be either forward-looking and do it yourself, or you can be reactionary, defensive, and have to do it because the other guys in your sector have done it. That extends to part of your question too.

Retail is indeed a great example. The brick and mortar retailers have clearly had a hard time with eCommerce-based retailers in the last few years and have been recently opening up APIs to try and give them another avenue to get back into the game.

Some of these retailers are getting back in the game relative to mobile. You have guys like Best Buy, Walgreens, Sears, and others with open APIs that try and be part of that mobile shopping experience of, "I need to go get an Xbox player this afternoon. Where is the cheapest one near my house? Oh, it's at Best Buy."

On the flip side, there’s the pain of having too much opportunity. We often hear stories of how many people opened up their API because they had to turn deals away because they just didn't have the bandwidth to deal with prospects that were coming in. But if they could point the prospect to an API, and that in turn could help us locate progress, partnership, and revenue, then that's a great thing. I think that's a key benefit of having an API, it can increase the funnel for the provider.

API Economist: John, it’s time for one of our favorite questions here at The API Economist. What are some of your favorite mobile apps? 

John Musser: Well, I certainly have a lot. Many of them are extensions and aren't necessarily mobile apps. Shazam has always been one of my favorite mobile apps. It's always seemed like magic to me: that I can to be anywhere, hold it up, and have it tell me what that piece of music is. It's been around for a couple years so it's no longer the novelty that it was a couple years ago, but it still gets a lot of use. That and SoundHound.

For services, I get a lot of mileage from a utility perspective out of something like TripIt. Every time I travel I use TripIt. It’s an app with point data from a lot of different sources to give me the mobile experience.

I use RunKeeper to keep track of my runs. Speaking of APIs, RunKeeper started life as a mobile app and now is a platform. It’s been so successful and they gathered so much data that they went from being an API consumer to being an API producer.

API Economist: John, thanks for your time!

John Musser: Thank you!