• by Tom Hespos , December 12, 2001

The idea of an audience-based model for buying and selling online advertising has occupied center stage at industry conferences and opinion columns lately. I've recently changed my opinion on how this model should function within our industry, and I wanted to give my $0.02 worth in this first article for The Online Spin.

There are quite a few industry opinion leaders calling for the end of the impression-based system we use now. They would like to replace it with a system that leverages the model that traditional media currently uses.

I think this would be particularly damaging to the industry.

The audience-based model calls for buying ads so that you pay for the audience of a site, not the impression. This would allow for compatibility between the traditional model and the interactive model, which would make it easier for traditional advertisers to integrate their media efforts.

Industry opinion leaders are getting excited over the effort of The New York Times Digital to implement a new "surround sessions" sales model. While I agree with many of them that the NYT model represents a step in the right direction, I also believe that the model needs some refining and that we should maintain the existing impression-based model so that those who wish to buy in this fashion can continue to do so.

One of the more exciting proposals I've seen in the wake of the NYTimes' announcement of its surround sessions comes from the Premium Network. I met up with Ken Margolis of the Premium Network last week at @d:tech, where he showed me an alternative pricing model for activity on his network that allowed for the purchase of unique users on his network within a given content category. The model also allows for frequency breakouts and is priced based on the number of unique users an advertiser wants to reach and the frequency with which the advertiser wants to reach them.

This is another step in the right direction toward compatibility between traditional and interactive buying models, but we still have a bit of work to do. The Premium Network currently allows for buying unique users, but it doesn't currently allow for targeting by common demographic audiences, which is the final piece of the puzzle with respect to complete compatibility.

After hearing Ken talk about his network's new model, I began to formulate some opinions about the movement toward an audience-based model and what is best for the industry. Clearly, we want to make it easier for traditionally-oriented advertisers to buy specific audiences online. This will help to generate incremental ad spending, which will help to stabilize our industry. At the same time, we can't sell the medium's targetability and trackability short. So we may want to continue to sell impressions while offering the audience-based model to those who want it, as both the NYTimes and the Premium Network have done.

But why limit the audience-based model to targeting only demographic audiences? With the Internet's enhanced targeting capabilities, we can beat traditional media at its own game.

Think for a second about how traditional media is plan

ned versus how it is bought. Planners come up with a planning target for their campaign, which is usually expressed as a demographic target with some psychographic or behavioral qualifiers added. A planning target for a packaged goods campaign might have a planning target consisting of moms with kids. But guess what? It's often difficult to buy media against moms with kids. Agencies that represent packaged goods companies might plan their media against this target, but buy against Women 25-39. This is a limitation of traditional media buying that is not present in interactive media.

There is no reason why planning targets and buying targets can't be completely synchronized when buying interactive media. Sites can target ads to moms with kids, based on information that might be collected when users register to use the site. If sites were to collect information from their users in this way, we can offer a buying model to traditionally-oriented clients that blows away the buying models they use in print, TV and radio.

I think that the interactive industry's next challenge is to implement the audience-based model with as robust a targeting mechanism as the medium is capable of, while continuing to offer the impression-based model for advertisers who are comfortable buying in that fashion. This is the quickest way toward increasing overall spending on interactive media and pulling the industry out of the slump it is currently in.