Want to launch a Facebook application that gains millions of users? Awesome, so do I! There are two challenges facing applications: (1) gaining critical mass and (2) retaining users. Andrew Chen has a great post about why apps eventually “jump the shark.” While his post is extremely technical, the lessons learned are valuable.
When the Facebook platform launched 9 months ago we witnessed the overnight growth of a few hundred applications. Many of these apps ended up with millions of users. Only a select few were able to retain those users though. Applications such as food fight or other Facebook classics found it difficult to retain their users. Ultimately they jumped the shark. Andrew has done an excellent job of outlining how and why this takes place on his blog post.
A select number of applications were able to actually retain their users and keep coming back. I had the opportunity to moderate a panel at the Graphing Social Patterns conference this week with a select group of individuals that have successfully retained their users. This lucky few had clever ideas early on. They were rewarded and now they have figured out successful models for retaining them.
Ultimately you can look at an application and tell if it will eventually jump the shark. Clever ideas are rewarded but frequently they go un-nurtured and result in the demise of an applications. Some ideas are meant to eventually jump the shark. JibJab media is one example of a company that is constantly innovating in order to retain their users. That is the nature of the viral business.
Facebook has created a mechanism in which clever ideas can be rewarded quickly but ultimately it is up to the users to figure out how to keep them there. For those that build applications, some of us will be forced to constantly chase after those clever ideas, others will have more sticky ones that retain users. Both models work but ultimately you must choose one or the other. Since Facebook cut back on the tools available to applications for going viral (primarily invites), both models have become increasingly difficult.
Check out Andrew’s post for more reading.