Liz Gannes' analysis on GigaOm . She notes that CPMs are decreasing amidst increasing inventory levels, and advertisers are reluctant to put their messages next to user-generated content. But there's more to the picture than that.

I see four additional issues that are contributing to this failing ecosystem of online video advertising. The first is a lack of a cohesive advertising infrastructure. Part of this is a technology issue. Advertisers, publishers and content creators are working with different video advertising platforms, formats, and systems to measure viewership and engagement. The result: Vendors and their agencies have to work much harder to evaluate and execute a campaign. In many cases, they are forced to limit themselves in terms of the type of advertising and the outlets that their messages are shown on.

Second, there is a huge diversity of video programming to choose from, much of it focused on niche topics or very focused audiences. In order to get a half-million eyeballs on a certain ad, agencies may have to divvy up their buy amongst a multitude of smaller programs. Extra work is required, and this increases the chance that they will have to accommodate the different technology platforms and measurement tools mentioned above. In addition, programs with small audiences may be excluded from consideration -- how many deep-pocketed vendors want to spend time and money on a program that has only a thousand dedicated viewers each week?

Third, advertising professionals in agencies and publishers' sales teams are not pushing aggressively into online video. Who can blame them? In addition to the problems listed above, there's still a lot of uncertainty associated with online video advertising. Everyone seems to be experimenting with features and formats, and new players are constantly entering the game. Others have shied away from video owing to fears of a recession. No wonder eMarketer has , and many advertisers and sales teams are sticking with models and formats that they are comfortable with.

Finally, one of the biggest obstacles to online video isn't on the advertising side. It's on the content creation side. Producing content is a very labor-intensive process, and video is a very demanding medium. Unless you've worked with video before, it's difficult to comprehend how expensive video can be. Yes, it's possible to take a low-budget approach to production with Flip cameras, YouTube, and other cheap or free DIY tools, but the amount of effort required to create a single piece of content is still relatively high compared to content based on text or still imagery. Sustaining any programming effort over time will suck up a great deal of time and money, and there is no guarantee that viewers -- or advertisers -- will flock to the program. Many ambitious technology publishers have learned a hard lesson trying to make advertising-supported online programming work, but have been forced to eventually cancel programs because the revenue is just not there.