The staggering amount of video YouTube processes confers serious infrastructure advantages. Achieving that level of scale provides critical technology operations insight. By the time everyone realized that video needed to become a core content offering, YouTube’s penetration and brand presence was insurmountably large. Their embed player, ugly as it was (and is), was (and remains) ubiquitous. Had YouTube rolled out a commercial video platform with
rich on-demand encoding services, a fully customizable embed player, HD or at least higher quality playback all sitting on a developer friendly App Engine-type infrastructure, they would have obviated the need for at least half a dozen related services from On2, Brightcove, thePlatform, Limelight and Akamai. They could have created the Amazon AWS equivalent in video overnight. It costs real money to serve video, and it ain’t easy, particularly when it’s done well. Video consumes orders of magnitude more bandwidth than other content types. As such, it presents a significant infrastructure business opportunity. However, most good companies opt to ‘monetize’ simply by marking up their costs, rather than playing shell games with loss-leaders and hoping to make up the difference with advertising. No amount of advertising can feasibly subsidize the cost of serving video both on-site and off-site, particularly at healthy operating margin levels.
Posted by Jonathan Marcus at May 10th, 2009 at 11:42 am
