A few days ago, I was talking with Pete Zafonte, owner of game and proxy designing company Etnofaz, about Facebook’s new activity-monitoring system of application development. His view: that activity is just as misleading a statistic as total number of users. Take the Vampire and Werewolve applications: while both encourage daily usage, both still rely entirely on drawing in new users without providing any sort of new functionality. Pete presented this (admittedly rough) metric for determining application value:

Clean interface
There are many well-designed Facebook applications that fail because users can’t understand quite how to use it. Every application should be clean enough for users to be able to figure things out. Look at Facebook’s own applications as guiding points: each presents current goings-on in their systems on the front page, but on the top-right of each application is a quick way to utilize each app: Write Note, Upload Photos, Share an Item. It’s easy, and it gets used a lot. If things are too hard to use, they won’t get used.

Object-oriented
By object-oriented, this guide refers to applications that are based around objects that the applications itself creates, not simply user profiles. This is another guideline that most current popular applications use: iLike’s application creates separate pages for individual artists with discussion walls and songs, and Flixter’s movie application creates pages for movies. Less favorable by this system would be add-on walls or systems like X Me, which go directly from user to user without any center meeting point in between: while applications like this are good for friend-to-friend interaction, there’s not much chance that users can extend their networks and find new people using applications like this.

Innovation (of new ideas or existing ideas)
This one’s really just an attack on some of the copycat applications that have been springing up. Users don’t need four extra walls, two more poking systems, and five music players. It’s usually a better idea for applications to develop new, unmarketed concepts in Facebook than it is for them to just rip off old ideas and hope things turn out okay.

Interoperates with Facebook or other Applications
Interoperating with Facebook right now means one of a few things: allowing for user tags (like Notes, Photos and Videos), allowing Wall Posts (Box.net’s Files application, for instance, makes sharing files via messaging a snap: yet another thing I’ve noticed is done more and more on Facebook and less via emailing), or through broadcasting useful Notifications (using Flixter’s app, for instance, I get quite a lot of nice feedback on ongoing movies).

There’s not much interapplication work going on right now, which is unfortunate. If applications ever figure out how to communicate back-and-forth with one another efficiently, the possibilities for a truly broad application network could be enormous.

Broad Audience (as much as possible at least)
If Facebook is promoting applications, it only makes sense that they promote ones that are the most useful to the most people. Facebook’s current system does this quite well, but this is on the list as a reminder that very efficient applications don’t make for the best applications if they’re very obscure.

The Etnofaz application guide will probably never be anything that Facebook could base a system off of: it’s too hard to measure through statistics, and it’s too slow to really encompass all the applications that are out there. It is, however, of enormous benefit to application developers: each of these guidelines is extremely useful for people trying to create the next big application out there.