Monday, August 06, 2007

Identity Crisis in the Land of Social Networks and Platforms

The Facebook platform debate continues to evolve with Dan Farber's recent piece on Facebook, social capitalists and open networks and Wired's very well thought out article Slap in the Facebook. The key question from my perspective is whether we consider the Internet as the platform or does Facebook or some other single entity come to dominate and become a platform. The history, and even the brief history of the internet, has examples of both - a platform owned by a single entity, and the internet itself as the platform with various platform players as parts of the whole:
  • EBay: Ebay is a platform and is a pretty closed one. They recently had the chutzpah to even take on Google by banning Google Checkout. The APIs and other interfaces to Ebay allow you to enhance the functionality but does not offer any interoperability - you cannot cross list items on Ebay and some other auction site, etc.
  • B2B: B2B Exchanges were an attempt to create a platform (remember Covisint) but eventually lost out to the Internet as the platform. Companies transact billions of dollars of business today on the B2B platform but they rely on protocols like RosettaNet and there is no single hub or platform that dominates.
  • Instant Messaging: IM is an interesting case study as it started out as platform islands (Yahoo!, MSN, AOL) but over time and sometimes grudgingly they have learned to play well with each other. IM is still not an open network in the sense that I cannot create a new IM service and seamlessly connect to these proprietary IM networks.
  • Email: Email is the ultimate open network. It has mostly worked great except the openness allows for spam and viruses to be spread using this platform. This security issue is a (valid?) excuse many platform players are using to keep their platforms closed.
What model is Facebook following?

So is Social Networking going the way of the EBay model, IM model or the Email model? Facebook today sits somewhere between the Ebaymodel and IM model. Under the Ebay model, Facebook does not enable to send messages back and forth to say MySpace - in fact, the messaging system could have been associated with an email address ( but is not. At the same time, unlike Ebay which blocked out Google Checkout, Facebook is allowing third-party applications to be shared and used in conjunction with its service - somewhat akin to the IM model. You still can't use your Facebook id to interact with someone that does not have a Facebook account.

The Identity Problem

Its been suggested several times that the lock in and lack of interoperability comes from the fact that the identity systems of Facebook and other services are not open and standards-based. If Facebook and others like Myspace all adopted the OpenID or equivalent identity system, it would be so much easier for users to leverage multiple services without worrying about whether they are built by Facebook, LinkedIn or MySpace. Irregular friend Dennis Howlett describes the conversation on Facebook use within the enterprise on his ZDNet blog post.

Dan Farber comments on this lack of interoperability:
Today, people are mostly content, experimenting with the more civilized walled gardens that aggregate information and friends and bank all the personal data and social capital. The revolution won’t happen until social capitalists realize that the capitalists–Facebook, Google, MySpace (News Corp.), etc.– shouldn’t have too much control over their digital lives.
Who will bell the cat?

I feel that the masses will not be the one's that change the status quo. It will be a game changer - a new Facebook or Google that will challenge the closed networks by offering a good enough service that is as good as MySpace or Facebook but is entirely open. In fact, Google could do this, and it would be much easier than you think. Here is what I would do if I were running Google social networks group (no, they haven't asked me):
  • Google has the email accounts of several million users.
  • Google could analyze my email messages to all users - this is where having stored all my emails helps - to determine my top 100 contacts. Repeat this for every user and you have created a social networking graph for all Google users and many non-Google users too.
  • Google could then instantiate GoogleBook (I own the copyright!) accounts for every Google user ready to be activated. All a user would have to do is select and unselect the suggested links and the account would be all ready to go. For non-Google users, a 'claim this' GoogleBook account would be created which they can claim by requesting an email be sent to their email address.
  • Google Groups - like functionality would be available for each user i.e., I can send messages to all my contacts, share calendar, files etc.
  • And since you are NOT required to ever create a gmail (Google) account with a new id, the users wouldn't be forced to create yet another dan.farber@gmail/cnet/yahoo/etc.
Whether Google or some other new player does this anytime soon is anybody's guess but many of us are getting sick and tired of creating multiple user id's, checking messages on multiple inboxes and accepting the same 75 friends on 10 different social networks. For now here is my personal solution to the social networking problem - if you have my gmail address and my blog address, that is all that you need to reach me, read about me, see my pictures, date me, send me fan letters and/or harass me.

Update: Dan Farber has posted a response to this post on ZDNet and the conversation continues.

Update: Dan Farber reports that Google is planning a foray into social networking. I expect them to mine my email etc. for helping create my network - as I mention above - let's see what comes out.


jitendra said...

Ashu, Check out SezWho...We try to address the identity issue by building context specific portable identities across site...

see an example at

Tish Grier said...


I'm just wondering if the Silicon Valley folks have thought that many people outside of its confines may not want one-stop social networking shopping. Perhaps people outside of the Silicon Valley,n businesses other than technology, don't feel any compelling need to join every single social networking application out there, and, subsequently, don't need to be bothered with mulitple social networking accounts. Personally, I prefer to keep my social networking accounts separate, as they represent separate parts of my life--Facebook for all kinds of "friends," my LinkedIn profile is for business (some overlap w/Facebook, but not totally), and my never-used MySpace page, which I had before Murdoch took it over, for checking up on one friend that was displaced due to Katrina.

And I would hesitate to give any more information about me to Google than it already has. The wisdom of that has yet to be tested by hackers.

Unknown said...

I think Tish makes a pretty good point, and one that is close to the impression I have about unified social networks, at least in the ideas I've seen proposed thus far.

a) openid is great for what it does, but the idea of every new service supporting openid seems to me like it would have a negative impact on the community of sites like flickr, for example. If I was some 13 year old myspace girl, and I could instantly become part of the flickr community without any barrier to entry, I think the community as a result would become less defined and less cohesive.

b) Most of the "portable identity" stuff doesn't do a great job making differentiations between communities. For some stuff (commenting reputation systems, for example), it makes sense to have one identity across different platforms. But dragging your contacts list around with you from network to network doesn't make a lot of sense, considering the wide spectrum of relationships inherent to such a list. The main use case for me here is more of a migration issue, which stuff like XFN could be helpful with: ok, myspace sucks, I'm packing up my people and going to play over at facebook. We'll hang out there until it becomes lame, then move somewhere else. This works because facebook and myspace are similar social spaces. In general, I'd agree with Tish that keeping social networking identities unique to certain communities is probably something that won't go away too soon.

Whew, that was probably one long run on thought...

Unknown said...

I agree that we need several social networks or role-based networking as opposed to one single network. In my role as a business person, I need to share and connect differently than in my role as a friend.
However, that is more a matter of providing multiple 'roles' rather than having to join multiple sites necessarily. The reason we join multiple sites is exactly because there is no role based capabilities available in Facebook/LinkedIn etc. We need finer-grained control.

Unknown said...


You make some excellent points and I agree that we don't want all our networks to have the same contacts. But as I replied above to Trish, this can be accomplished by sites supporting the concept of roles and allowing me to manage relationships. For example, LinkedIn only supports one relationship "is connected to" i.e., Anshu is connected to Bill etc.
They could allow me to color my contacts by relationship type:
-Business associate
And then allow me to manage and share my network based on roles.

The OpenID is just one tool that we can use to enable cross platform interoperability. Its definitely not the full answer.

You make some good points here though - let's keep the conversation going.

Anonymous said...

I don't entrust my online identity to one entity. And I minimize my interaction with google. I block incoming gmail at the MTA. Google is not your friend. And you'd be wise to write an article/reflection on the overlooked agency in the trust model..