by Charlene Li
Google and a slew of partners announced the formation of OpenSocial (URL will be live late Thursday), which Google's Joe Kraus, in a briefing earlier this week with me, described as "a common set of API's for building social applications across multiple sites." There are three specific sets of APIs that tap into 1) member profiles, 2) the social graph, and 3) member activities.
The idea is that developers would have to learn only one set of social app APIs and create apps that will work on any partner's platform. Moreover, the OpenSocial API is written in HTML and javascript (and supports Flash), compared to Facebook's proprietary API. Initial partners in the new API include social networking sites Friendster, hi5, Ning, Orkut, Plaxo, and Viadeo, as well as application companies Oracle and Salesforce.com.
Note that there are some major players not included, namely Facebook and MySpace. Also missing are Microsoft (aligned with Facebook) and Yahoo!, as well as other smaller social networking sites like Bebo and Piczo.
Here's my take on what this announcement means:
* Facebook isn't threatened -- for now. Application developers are going to go to where the heat is, and that heat is red hot at Facebook. They have not only a lead with 6,000 social apps already on their platform, but also close to 50 million users actively using those applications (Hitwise reports that Facebook traffic is 9X the OpenSocial coalition, but this doesn't include international traffic). Add on the third leg of the social app stool -- monetization, which Facebook is set to announce Nov. 6th -- and you have a developer's dream. Any developer worth his/her salt is developing on the Facebook platform, trying to figure what works, what doesn't. And because of this head start, developers will still develop for Facebook FIRST before developing for OpenSocial. It's similar to search engine optimization -- most Webmasters optimize their sites for Google's bots first, and then worry about the other search engines.
However, in the long term, Facebook's ability to dictate how social apps are built will deteriorate under two scenarios. First, closed platforms put a tremendous burden on developers, so the benefit of a vibrant community has to be enough to justify that extra effort. Facebook has tremendous growth and vibrancy, but it will have to sustain that in the face of increased competition. Second, as social applications will spread beyond social networks (see below for more details), Facebook will be less and less able to act as the primary destination for all social application usage. For example, Facebook may still require that the only way someone can read book reviews from friends is on Facebook via the Amazon social app, where as Amazon may be able to show OpenSocial member book reviews on its site.
* OpenSocial gives the partners a fighting chance. Now just because OpenSocial won't displace Facebook in the short term doesn't mean that it's a failure. In fact, it's just the opposite. Without OpenSocial, none of the partners would have had a chance of capturing developer time and attention.
* Developers win big time. Let's say you're a wannabe social app developer -- you're dancing in the cubicle because you can now get lots and lots of distribution on new sites without expending a huge amount of effort. This is especially important for developers who haven't made it big on Facebook, which is dominated by early movers like Slide and RockYou. Watch for the smartest, most aggressive developers to move over to the OpenSocial platform, eager to create apps that will gain early, viral traction in these new networks.
* Social apps will go beyond social networks. Note that Oracle and Salesforce.com are also partners. They have a strong interest in "socializing" their applications -- I've written about applications like FaceForce that pull profile data into Salesforce.com. This opens up a whole other space for OpenSocial, namely any Saas or online site that would benefit from social information. Examples would include recruitment sites like Monster.com or CareerBuilder and dating sites like Match.com.
* Things will get interesting when MySpace opens its platform. MySpace announced at Web 2.0 plans to open up its platform, but no details have been forthcoming. If MySpace decides to remain independent and not work with OpenSocial, developers will then have three APIs for social apps to learn,further disadvantaging OpenSocial. But if MySpace is smart and can set aside its competitive impulses, it has more to gain by supporting OpenSocial as a default standard and stranding Facebook all by itself.
* This isn't truly "opening" up social networks. OpenSocial is a set of APIs for developers to create social apps ON platforms that leverage data within each specific platform. One of the biggest complaints about Facebook today is that it's a "walled garden", meaning that all activities that leverage its social graph have to take place on Facebook itself. As I wrote above, imagine the utility of being able to read your friends' reviews in Amazon, or having a social graph of a contact appear in Salesforce.com applications. We're a long, long way from that happening anytime soon.
Lastly, I think OpenSocial will re-energize social networks as they broaden the activities their members can do on their sites. In particular, I'm looking forward to moving beyond the typical fun yet frivolous apps like food throwing developed by 20-something developers. I'm going to closely watch LinkedIn, as collaboration and expertise location applications built on top of its professional business networking social graph will make the site more relevant to me.
I'm interested in your take, especially if you're a developer -- will OpenSocial make a difference in how social apps are created?
Other relevant stories from:
New York Times
TechCrunch
ZDNet
ZDNet's Garet Rodgers on developing apps on Open Social "Google's OpenSocial platform is great!"
Ning backer Marc Andreessen on Open Social
Tags: Facebook, OpenSocial, Google, social apps, Forrester Research, Groundswell
http://blogs.forrester.com/charleneli/2007/10/goog...