ABC Experiment

I watched an episode of 'Lost' last night on ABC.com just to check out the experience (as for the show, I was little lost, since I'm not a regular). I thought it was a very good implementation. The initial navigation 'arc' from which you select a show is a little awkward, but not bad, and the player is good and clean and the video quality was very good -- really quite watchable in full screen (or expanded screen, more appropriately). The ad experience (its a little disturbing that my work has now caused me to watch online video with greater interest in the ads than the programming) is just OK. They are doing what MTV and others are doing with long form content which is multi asset chaptering -- i.e. using the existing program chunks and basically just putting a post roll video ad break behind them. That's fine when its one ad, one sponsor, small audience, low bar. But I think the right way to do it is single asset chaptering where a parallel data file signifies opportunities for ads and ads are dynamically inserted (or not) depending on a wide variety of targeting and business rules. At ingest one can identify a larger range of opportunities for ad opprtunities (not to be confused w/ ad breaks), tag content in a number of valuable ways and make on the fly decisions to insert advertising or branding elements, go to a break or not, depending on what and who is in advertising mix and what their objectives are. Better user experience, better advertiser experience.