One of the least considered aspects of Social Networking sites around the build VS buy decision.

Social Networks are a process, not a product. When you're building a Social Networking application, your actually looking to enguage customers - all of your value comes from repeat interaction with eyeballs. When you're considering building a Social Networking application, you really need to be cognicent of the value of feature release.

Despite what vendors say, it's still a bit of a mystery

One of the unexplainable aspects of Social Networking sites is why some products, sometimes nearly identical to their competitors, fail, while others succeed. One of the aspects that contributes to success or failure is actually the release of features that relieve customer pain or add customer value. Every time there is a release, you're incubating the activities that drive the core value of the network - talk. You're getting mentioned, you're seeing discussion about the features, and you're actually driving the corresponding value of the play up.

Driven by peaks and valleys

Most Social Network site traffic works by peaking, and then coming back to a level a little higher than the previous level. What causes this? Primariliy, it's the community reacting to something. And quite often, it's the release of some cool new feature or some amazing new tool that inspires various developer communities to take a look.

In fact, it turns out that the process of adding, amending, and creating new features for Social Networking sites is one of the major drivers of their success. Products that ship in a complete state experience little of this effect.

The value of being open: unexpected traffic

In additiona, due to the fact that other people's products can be build upon your service, it actually turns out that the "network effect" or third party product releases - even if only tertiarly related to your product or service - can have a massive impact on your traffic, and therefore core value.

Remember, the game is eyeballs, not control.

Increment to success: The new model is iterative

So, when you're looking at the build VS buy decision, it's really important to consider the implications of ongoing, incremental releases. I often advise people building Social Networking sites to hold back features from the initital release. The rule of thumb is "ship the core value" - then release your new "cool features" one at a time, marketing the heck out of them around each release. After all, what gets more attention in a world with a short attention span - a project that ships something amazing or a project that ships something amazing every week?

The development process and approach itself can drive traffic

Incremental feature release encourages repeat visits. It's the same reason people keep going back to online games, checking to see what new weapon they're going to get, or what new bell and whistle they're going to get next. Take advantage of your user's desire to discover new things, and create as many touch-points as possible, but don't give them too much at once as they might get overwhelmed. Let a feature sink in for a bit before you move on to the next thing.

Please users and win, fail to please users...

Remember the most important thing you can do in a Social Networking play is suprise and delight your users, or remove the major pain points identified by your users or through user testing. Remove the barriers to adoption, and never, ever let your site become stagnant or disappointing to your users.

The disappointment ("I wish they'd get it") is what kills sites.