The Directionality of Social Network Platforms
Posted 10/18/2007 10:49:00 AM |
Why does a web service implement an API? There are many reasons - one of the most consistent is that a service seeks to increase value by sharing its content beyond its boundaries. For example, Twitter leverages global programming and design skill to put their simple product into people's hands on their terms. Google Maps, as another example, exports its data to enrich other applications, producing a transformative functionality.
And what governs a firm's decision to produce said API's? It's simple economics; API's are simply new routes to pre-deployed functionalities, and whatever costs the firm incurs (generally minimally-variable costs in terms of infrastructure) to provide the API is recapitalized in terms of market penetration, novel functionalities and user satisfaction. Thinking outside the firm, what drives outside adoption of API's? For the external firm, the adoption of the API requires learning a new skill, data integration, programming, testing - not an unbearable cost to a mom-and-pop, but clearly not a simple decision for larger shops. Therefore, a promise that what one gets in return for time and resources spent figures rationally into the decision.
Lets wander back to Google Maps for a second. Putting aside the fact that Google Maps API is dead simple, perhaps the best in the Web 2.0 sphere, what does a firm know they are going to get by integrating? Answer: A fantastic, market-leading mapping functionality that will produce value almost instantly. In essence, there is a give-and-get with API's; firms will expend resources if they know they are going to have value driven back into their products. This is a rational actor assumption. Is it irrational for firms to design for social network platforms?
Before we declare social network platforms a giant house of cards, let's define the space a little. When one designs for Facebook, they are designing a tool to be run in a third-party space. That is, while the business logic of an application resides on the firm's servers, the experience is entirely in Facebook. So what does the firm get by designing for Facebook? If the application becomes popular, one can expect name recognition, acclaim, an appearance on TechCrunch. But what about the other 99.9%? It is clearly a gamble with little guaranteed return.
This brings us to the idea of directionality in API's. For a relationship between two companies to be successful, it must be mutually beneficial. With regards to the Facebook Platform, Facebook clearly benefits - the 5500 or so applications developers have created makes Facebook better, at minimal cost to Facebook. But what do third-party developers get, other than hopes and dreams that they might sell their app to Slide one day?
Third parties expend effort in designing for the Facebook Platform. But what is Facebook giving up? I'd argue that Facebook is giving up nothing. By implementing draconian terms of service, the API is essentially unidirectional. If you look at what is "storable" on the third party's side, one sees that Facebook provides third parties no valuable data. And of course the users can't override Facebook - you don't own your data, your interests, your friend list or your photos, so you can't provide that information to third-parties via the API.
My profile data, my "social graph" - all of this data could go a long way towards enriching other applications. Since the Facebook API only flows into the service, not out of it, third parties can't leverage this value - they must simply bet on their applications being a hit in the service. Looking over the most popular apps, what do we see? A set of fluffy, fun, particularly useless applications - Kara Swisher echoes. Stepping back a bit, I realize these apps are valuable - people like them, they're fun - but the ecosystem needs more than fluff, especially if we're going to start talking about the "social operating system."
A few weeks ago Anil Dash wrote about another unidirectional system, declaring that Facebook will eventually blend into the web at large. I agree, and I propose this will occur when Facebook realizes that it must bidirectionalize its API. Although interest in the Platform may have peaked, we're simply not going to see truly rich applications (that require substantial third-party investment) until Facebook starts giving back to those who use the API. Looking at limited uptake by presidential candidates, one can only imagine the incentive these candidates would have to create awesome applications if Facebook bidirectionalized the API.
Right now, Facebook can get away with their stance due to their advantage in the marketplace. As Myspace chases their heels, and third parties realize the limited value in a unidirectional API, perhaps this will change soon. If nothing, Facebook should support identity rights - let me control my data and decide where it goes. If they don't, perhaps the market will chart a new course for them.
Permalink |
|
to this post
View blog reactions | Post to
1 Comments: (Post a Comment)
-
At October 21, 2007 6:52 PM,
Bertil Hatt said...
-
It sounds a little bit like gambling: the odds are fair, but the house has pockets so deep that it cannot loose — while the players, they play until they are bankrupt, which happens soon enough. Facebook has an essential asset, so it can force the others into something so specific that. . . Facebook only value, the social contract it is based on, can be threaten by the fear of that imbalance.
Why anyone else then a Whale big enough to be able to make the house loose would enter a Casino? I dunno—but Facebook might want to investigate. Maybe it's that human incapability to use really thin probabilities again.
