Yes there is money in API’s

Dameon has written an excellent article asking “Is There Money in Voice API’s” for Gigaom. This article has some of the best comments I have seen since long time. (Check my comments as well). I couldn’t resist talking about this topic. I agree with dameon that startups cannot just open up Voice API’s and call that as a business strategy. But I see lot of potential in these API’s. In general, API’s regardless or whether it is voice, messaging, video, music etc has a great potential to make money. The question really is who gets the Pot of Gold. I see a great future for mashup applications that can render data from different platforms filtered and customized as per user’s taste.

The way I look at API is; it’s just an additional avenue for companies to make money. For e.g., twitter opened up their platform with API’s that can be used to build applications on top of it. We saw tones of apps that were built and some (summize, twirl etc) of them were bought. So what did these apps do to twitter; 1) they increased the usage of twitter, 2) Developers built innovative apps that pulled data from different third-party servers. 3) Twitter branding. Similarly if you look at other products like YouTube, Last.fm, Google Maps, Flickr, Yahoo mail, Friendfeed etc. All these products provided API’s using which developers could build applications. Here is some statistics from programmableweb ,
There are currently 3209 mashups with 822 API’s.

There are some folks that are pure API providers like ribbit, who are into the platform play. The model here is to provide API’s for developers to build standalone applications. What is missing here is the branding. For e.g., I can build a mashup application that can pull up Youtube video and a matching flickr picture. There is already a brand called YouTube and Flickr. However, in the case of ribbit, they are just a pure platform player. So the branding is missing here. In addition, ribbit business model solely depends on the developers. It makes money only if the developers succeed in building compelling apps.

Another concern that most of the folks raised was standardization of these API’s. I couldn’t agree more with these folks. I guess there should be some kind of standards that needs to be followed with these API’s. If not, we will see the repeat of SILO models, where every API’s shall be proprietary and you might end up learning 101 different languages.

I still wonder what the deal with operators is; I mean they are ones that should have been playing a bigger role in pure platform play. With the infrastructure in place, all they need is glue that bridges the platform and the API. Isn’t this the IMS role?