May 2010
Sun Mon Tue Wed Thu Fri Sat
« Apr   Jun »

Month May 2010

Quick Thoughts on Swirl

An Unexpected Find

One thing about working for a Fortune 100 company is that there is going to be swirl, red tape and politics. Granted, this is true for smaller companies but the stakes are smaller and the separation between disciplines are grayer. The bigger the company structure the deeper the divisions between teams, it’s just the economy of scale rather then any malevolent corporate doing.

Well, for the last few months I’ve been working on the next generation of one of the Interwebs more trafficked sites. The idea was for my team to create the ideal “concept car” that would inform the current and future production versions. Our concept was more prêt-à-porter (ready to wear) then couture, still it needs about a year of back-end catch up before the vision can be realized. During the conceptualizing phase of the project we’ve tried to connect with as many of the teams that would be affected by the new site, find out what they needed, their concerns and when possible give them a glimpse at what we were working on. Our goal was to be inclusive and make sure we weighed the often divergent requirements we received. It’s a been a great experience and overall it’s moving alone surprisingly smooth considering all the pieces we’re juggling to keep everyone on the same page and this is where the swirl comes in.

We present a stage of the concept by walking the teams through the thinking and aim to clarify any questions or concerns that come up. Still, once we present a concept what becomes of it through the interpretations of the various teams is beyond our control. Team A translates it based on their needs and fears with no knowledge of Team B, who are doing the same thing. Add to this layers of management, where the concept is distilled and re-presented in a telephone game style of communication. Now add to that politics and territorial pissing contests. Give it a week and there’s chaos. Assumptions and fears swirl through the departments like it was a virus. The next few days are a mix of moving forward and attempts to reign in the swirl created by our latest concept presentation.

Swirl is the worst part of the process. As a team, we’d gladly walk the other teams through our concept but the reality is we can’t. They’ve got projects of their own to work on and we have an aggressive deadline in which to move from concept to production. Swirl is guaranteed, at least as far as I see it.

Getting to know Facebook’s Open Graph

Last week Facebook announced Open Graph, a way for external sites to connect Facebook’s users and services. Open graph takes what was previously called Facebook Connect a bit further. First, it’s based on using OAuth rather then Facebook’s proprietary system. Secondly, it allows tighter integration with Facebook’s data as well as allowing 3rd parties to locally retain user data for more then 24 hours. This is a huge perk to 3rd party developers; on the flip side this is a huge privacy concern for members. Other additions include the ability for any site to embed some Facebook functionality without requiring OAuth or compromising user privacy, this is done through iFrames.

It’s this last one that initially will be Facebook’s biggest win. It includes the ability to add a “like” button to any page with a simple line of JavaScript. For external sites, like this one, the addition of a “like” button allows for a simple way for readers to share the site/page with their (the readers) friends on Facebook. Much like the Digg, Buzz Up, “post to Twitter” and other social buttons previously available. The big difference is access to Facebook’s membership numbers, especially since Facebook users encompasses a broader spectrum of the masses then most of the other services. Granted the “like” button is little more then a marketing/promotional tool for the sites that use it. For Facebook, on the other hand, it provides endless user preference data, which can be used for better ad targeting or be sold to 3rd parties. It also positions Facebook as the go to source trending information.

In addition to the “like” button, Facebook’s Developer’s guide offers a number of other widgets via iFrame/JavaScript for 3rd party sites to connect with Facebook. One widget offers a view of what other pages on the site your friends (and the world at large) “liked.” Almost instantly there was, a site displaying multiple instances of this widget listing what friends “liked” on some of the Internet’s bigger sites. All this is done without ever directly knowing who you are as each instance of the widget is actually an iFrame containing Facebook. I say “directly” because the data is readable via JavaScript post rendering. Facebook also offers a recommendations engine based on all the data collected that 3rd parties could implement. Though I need to learn more about this one myself before I can explain its particulars here.

One thing Facebook is requesting as part of the “like” button spec is the addition of meta data about the page/site that is hosting the button. The more accurate the info included the better you’ll be found within the Facebook universe. This reminds me of the early years of search engines where they relied on the honesty of the poster’s meta data for the integrity of their search algorithms. Regardless of the potential for hacking the system, Facebook has a list of requested meta data fields to be associated with the “like” button. The “type” field is potentially linked to the content listed in the user’s profile. For example, if you “like” a movie on IMDB, it can potentially be added to your profile’s list of favorite movies. Though I believe this type of connection to user data is reserved for OAuth connected clients rather then the JavaScript based “like” buttons.

The OAuth connection allows for more access to user data then ever before. OAuth now has access to profile data and the ability to locally store Facebook user data for periods longer then 24 hours. This empowers developers with the ability to parse through all the user data and make analytical connections that were previously impossible. For example if User X allows access and a week later User Y allows access, if User X is friends with User Y the data is now available to make this connection and any others that come along with the increased dataset. Additionally as an admin for pages that were “liked” you can push page updates to those users.

Overall, Facebook’s Open Graph looks like it’s worth using even if it’s only for the simple marketing benefits by adding a “like” button. Granted the real power lies in the OAuth integration. The biggest winner in all this is Facebook, as all these services places Facebook at the center of it all. As Facebook’s gravity increases they can always switch to a pay system so I don’t suggest relying solely on Facebook for the future of your site/service, but until then no reason not to take advantage of all that Open Graph has to offer.

Facebook: Open Graph
Facebook’s Developer’s guide