Web video is not a new medium, but it is new enough to not yet be refined. There is no template or formula that is a standard as of yet (maybe there never will be). While in the process of editing Season 1 of DOJ (and re-editing and screening and re-editing again and again...and trimming and paring...) I've learned a couple of things about elements that may work and elements that definitely do not.
Here are a couple of things I've learned thus far>>>
~Keep it short, punchy, and moving.
After screening an episode of DOJ featuring a fairly long conversation scene (not "My Dinner with Andre'" long ((a favorite of mine)) but long, nonetheless) recently to big laughs and a few kind words, on the big screen, I screened it on the smallest screen for a select group. The reactions were markedly different.
While the dialogue worked when the characters loomed over the audience and the speakers boomed, it felt a touch long when the viewers loomed over my laptop and the speakers chirped.
I must admit that when I began producing web based video content, applying lessons learned working on films and in TV seemed the natural route.
This was presumptive at best and dead wrong at worst.
Not to say that story or production quality should be sacrificed, but running lean is clutch.
When a video is contending with whatever else may be bouncing, blinking or bleeping on one's browser or desktop, that video better be damn interesting and fast enough to allow the user to return to work duties or Perez Hilton before they lose interest.
There is most definitely a learned psychological pigeonholing of web video. People hear "webisode" and they think "This here cat running around a toilet bowl better make me chortle in less than two minutes or back to FaceBook-stalking my ex-girlfriend I go."
When someone straps in to watch a sitcom on Hulu or a film on their machine, however, they are in for the long haul. Until the semantics change, it seems like original web based video content will have to run lean.