In a day and age of software accessed via web browsers, the fundamental development and design models of the desktop age are being turned on their heads. Heavyweight design, development, release and maintenance software processes have been replaced with lightweight, agile and flexible methods of supplying software and services that meet new customer expectations. In O’Reilly’s definition of Web 2.0, lightweight programming is about syndication of data and service and also the remixability and extensibility of functions and services to the benefit of the user and the provider.
One product which follows O’Reilly’s guides and which can assist companies during their start-up phase (where cheap is always best) is Facebook Connect. Facebook Connect is Facebook’s authentication and user information management platform which the company provides as a convenience to sites and services which need a login and user information management system.
Facebook Connect functionality is provided through an Application Programming Interface (API) which Facebook manages and maintains. This reduces overheads for the companies using the Connect service as another party is managing the technical aspects of interfacing with Facebook’s internal data stores and functionality and making necessary updates, while the user simply implements Connect and things just work. The benefits provided to organisations implementing Connect expand when the service is used as a login system for these organizations. Their costs of managing secure login services and user data management plummet, and Facebook benefits from extended reuse of their data.
As Facebook is a massive web service, the company already addresses one of the core requirements of the lightweight development/doing more with less design pattern by leveraging their existing cloud infrastructure. As Facebook manages all data centre work and all data management activities, organizations implementing the Connect service do not have to worry about managing any of these commoditised activities. Facebook can easily manage peaks in requests if something on Facebook Platform goes viral.
One final thing Facebook does well with Connect is the syndication of their business model through Connect. Facebook builds it’s business on advertising, which depends on data, specifically data created and maintained by individual users. By offering Connect, Facebook essentially expands their data collection reach outside of the immediate organisation and onto any site that uses the Connect service (especially through the Like button). By collecting more data, Facebook has more data to leverage to their strategic advantage in serving ads effectively.
The only issue Facebook has with respect to doing more with less is monetisation of the Connect service. The service itself does not produce any direct income for Facebook, which solely relies on the extra data gained through Connect implementations to deliver more targeted ads within Facebook. If Facebook wished to expand their ad network as part of Connect, managing advertising on external sites and services would be a massive overhead just for the Connect service and would likely require a separate part of the Facebook Platform to be developed.
Overall, Facebook Connect is an interesting example of a scalable, lightweight web 2.0 technology that has a lot of potential for both Facebook and the users of Connect. It will be interesting to see how services like Connect (especially Google’s +1 and Buzz services) change and compete in is area in the future.
Hi Trent,
Fantastic post! I completely agree with everything you said (as you would have seen by my blog).
Facebook Connect is a very interesting example of scaleable Web 2.0 technology. I liked how you were able to integrate O’Riley’s Web 2.0 Patterns into your discussion. Especially the collection of data from extrnal sources because this is SO SO important in the success of businesses.
Thanks,
Anthony Deacon
http://anthonydeacon.wordpress.com/blog/
Hi Trent (I always want to call you Heath as your first name!),
Great post – I learned a lot about Facebook Connect that I didn’t know before! I think it’s a great concept and a wonderful example of scalability and lightweight models.
What I did find really interesting is that the service itself does not produce any direct income for Facebook. I think it’s interesting that they rely on the extra data gained – I guess I’ve completely underestimated the value of user data to companies!
Mindy
I hadn’t realy thought about Facebook connect in this way before. Didn’t really think much about it all actually. It is an excellent example of this pattern.
I liked how you used a well established big player as your example, where as most others have blogged about smaller ( or newer) entities. Doing this had made me think differently about what this pattern actually means. Well a just a bit. thanks for that.
For me this topic was a hard one to find an example for, but Facebook connect is a fantastic example that never crossed my mind! To be honest I had never really considered the model for collecting data (both internal and external to facebook) and your post made me want to learn more. Interesting that such an integral part of facebooks revenue doesn’t directly produce income.
Good example from the other point of view: There you see that there have to be companies on the other side offering services to be leveraged by other companies following a lightweight approach. Interestingly, the game turns out to be a win-win-situation, i.e. there is no player exploiting another one to be cost-effective.
wow this is really a very good example for this pattern! I didnt know tat clicking on the ‘like’ button had a deeper meaning behind it. This kind of implementation creates more ads for facebook which in turns produce more income for facebook. Well written blog!
Excellent post and for sure a great example of the weekly pattern.
But I have to admit I don’t like the idea. I was recently on a website which just offered this Facebook authentication and I decided not to login. I feel somehow followed by Facebook, it seems to be everywhere and I have no idea what information is interchanged when I login on a third party website. Do you? In my opinion such powerful “everywhere authentication” should be provided by a more trustful company, in best case by a non-profit organisation.
You raise some good points. Of course security, and especially privacy, are incredibly important here. This could be one of the downfalls for companies implementing Facebook Connect as their sole login system. Consumers want choice in some cases. What it comes down to from my perspective is if the company using Facebook Connect really has the resources to afford the overhead of user management. Facebook does address the issue of trust over information exchanged and made accessible by using Facebook Connect by asking the user to authorize each new application, which is accompanied by a specific statement of what information is accessible to said app. An alternative, as you suggest, could be OpenID, which Anthony (see above) used as an example for this pattern.