So let's get the premise of this piece out in the open and then go about explaining how I arrived at this conclusion: The growth of video content online and its various distribution opportunities make any particular Web site almost irrelevant. This means that the way that pre-roll advertising is bought and sold must be radically changed.

Let us take a stab at discussing how "traditional" online advertising has been bought. First, one used to target Web sites that fit the demographics of their target market. Once upon a time, this made sense. Many agencies still want to buy advertising based upon a Web site's "brand" or demographics. That is why a lot of media plans obsess about Web sites, with agencies often positioned as the intermediary or gatekeeper to select the "best" sites for their clients.

But Google's AdSense seemed to suggest that the CONTENT on a site was more important than any specific Web site's "brand." So AdSense would serve contextual ads that corresponded with the content of the Web page. Advertisers really did not know where their ads were appearing, only that their ads were appearing within the right context or content. This was the move beyond targeting Web sites to targeting content.

The theory is that if you are reading about diabetes on XYZ.com, you are as interested in diabetes as you would be if you were reading the same article on Diabetes.com. Contextual ads geared for diabetics could be served in both places. The Web site brand is not as important as the content within which the ad is placed.

All portals understand this. Yahoo has different ads for Yahoo Finance than it does for Yahoo Sports. It is still Yahoo, but the content of the Web pages is matched with the right advertisers.

But with traditional media agencies, the individual Web site is still primary. These agencies seem to be missing what is really going on because of self-interest. Remember, being the gatekeeper is how they often justify themselves, so they exaggerate the importance of Web site selection.

So we see that content, even if only text, is king. But this theory is about to be kicked into full gear with video content. With video content poised to become the most desirable content to which advertisements can attach, how important is the selection of a particular Web site? Pre-roll advertising will attach to video content wherever it is. It could be on a video iPod. The content can be on a well-branded Web site, or it can be on a Web site that gets only one thousand page views a month. The video content is the packet or brand that is important, not the Web site or even device.

When a user initiates the view of video content, the audience is self-selected. The audience for "Seinfeld" clips is the audience that an advertiser covets--regardless of what Web site or device it is on. This fact is missed by most media agencies right now. If they do not understand the new dynamics, they will make themselves obsolete.

With new video content distribution models, we are seeing the same content being distributed on multiple sites. It is similar to syndication. You can get Reuters Video News on hundreds of Web sites.

This leads to the video content being the most important part of the advertising equation in the future. Advertisers will not care which Web site is the carrier, they will care what video content will accompany their spot.

This commentary is insightful. I recommend it to others.