• by October 4, 2000

By Jim Meskauskas

I've been harping a lot lately about pay-for-performance advertising; with clients, colleagues, and in print. I have made no secret of my enthusiasm for pay-for-performance advertising as it pertains to the online media space. As it stands, pay-for-performance advertising has really always been less about advertising - traditional branding and awareness - and more about an ancillary sales channel.

The beauty of the web is its accountability. I can track not only post-click activity but also post-exposure activity. Like never before I can get a value of granularity as it pertains to the quality and quantity of transactions as a result of my advertising.

The old saying used to be that I knew 50% of my advertising wasn't working, I just didn't know which half. The web tells me which half is working. What this has done has moved advertising from an actual world of correlative relationships between the messaging and it's effect to a potential world of causal relationships between the messaging and it's effect. And, as I said, this has created the conditions for advertising being used as a direct, albeit ancillary, sales channel.

That, in turn, has necessitated pay-for-performance pricing structures.

This position of mine, which is quite popular among advertisers and buyers alike, has not endeared me to my friends on the publisher side. It is no secret as to why: in their eyes I am compromising the integrity of their inventory, reducing online ad inventory to a chattel commodity and not recognizing the value of, say, environmental associations or audience quality.

So, let me set the record straight with everyone.

I do LOVE pay for performance advertising and I think that the advertising industry, both on and offline, is going to see more of it rather than less of it. But I do not think that it is appropriate in every instance, and I'll tell you why.

1. Use of advertising as a sales channel is only ONE way to affect the bottom line of an advertiser. Regardless of our ability to use the web and read causal relationships between our advertising and its effects, one cannot dismiss the fact that there is something to be said for advertising that does not result immediately in an action. Otherwise, everyone would be doing direct response television, radio, and print and you would never see Coca-Cola ads in a network program. When doing a buy, get firm confirmation from the client as to whether or not immediate impact on "moving widgets" is the goal.

2. Though, as I said before, I'm an advocate of pay-for-performance, and I still think it's the right thing to do in every campaign, the buy can be like Heisenberg's Uncertainty principle. I can tell you the velocity of an electron, but not the location; or I can tell you the location and not the velocity. Here, we can guarantee the advertiser the CPA (cost-per-action), but not the volume. Or we can go big with CPC and CPM and better guarantee them volume bu