• by Ari Rosenberg , Op-Ed Contributor, September 10, 2004

Last week, the problem in advertising on the Web was identified as relating the medium to a false pretense: print. This week, the column will address a manner to fix the system.

Of course the model in place does work, but the benefits of this model overwhelmingly favor the buyer. Who wouldn't want this kind of control over their purchases? So why are publishers so slow to react or even recognize this flaw?

Because budgets continue to increase for Internet advertising at a significant clip, year-over-year, revenue will continue to increase for major online publishers regardless of this flaw. But the saying "if it ain't broke don't fix it" does not apply here, because the publishers are leaving money on the table. Too much money, and the table may collapse.

So how can we fix the flaw?

Here is a hint; the Internet as an advertising medium does not resemble print or TV as closely as it resembles another medium. What was the Internet called before it became the Internet? The Information Superhighway. Can you see the road we are heading down?

The medium Internet advertising most closely resembles is actually outdoor advertising. Think about it. Users "travel" from one "destination" to another in search of content "stops" that fulfills a particular need. And as the user travels along the Web, billboard-like ads are displayed. How does a billboard company and an online publisher define the size of the audience that each respectively delivers to the door step of advertisers? By the size of their traffic.

This model can easily be adapted to online and the results are equally positive for the online publisher as they are for the advertiser. I know this to be true because this is exactly how we sold our inventory at IGN.com.

And here are the five steps we took to transform our business model from a flawed one, to a model that simply fit better.

1. We identified the areas of the site that made the most sense to sell a fixed position for a fixed period of time. This meant we would not sell rotating ad impressions generated with each page view, but rather the same advertiser would occupy the ad position for a defined period of time which we decided should be one full week.

2. We determined the front page of the site, as well as the front page of our seven most popular channels would be the most appropriate place to sell these fixed positioned ads. The ad unit we ended up selling was a banner/skyscraper combination served simultaneously.

3. To create more value for the advertiser, we stripped out all other ad positions on these front pages, so the advertiser would own these pages exclusively.

4. We then ran weeks of testing to determine an accurate amount of page view impressions we could estimate over a fixed period of one week. This "estimate" in essence, acted like a rate base advertisers could expect us to meet (we were conservative with this estimate so we always exceeded "rate base" as opposed to under delivering).

5. This weekly estimate directed us to the rate we would charge advertisers. Based on a market acceptable CPM of $15, we were able to generate a cost per entry an advertiser would have to pay based on the estimated traffic a front page would generate for one week. So for example, if the front page of IGN.com delivered on average 2,000,000 page views a week in a given quarter, we multiplied that by the CPM of $15 to come up with a rate of $30,000.

More to come next week on how this system works out for advertisers.