Spinning Remnant into Premium With Better Targeting


spingold_smallADOTAS – By treating premium ads as forward contracts, agencies are reserving first-impression ad buys for pumping their creatives through the pipeline by a specified time. Remnant ads, on the other hand, are less sensitive to time and placement, and typically go through exchanges.

The ad demand dynamic looks like this:


Premium demand for ads is by and large more vertical, which means that as prices for ads rise, the quantity of ads stays the same.

Remnant demand is not nearly as steep and is also inside the premium ad demand curve. Why? Because the quantity of ads demanded for any given price will always be less than the premium category.

Combining these two sides of the ad demand looks like this:


Notice the change from the premium category to the aggregate? This is the value of converting remnant to premium ads.

The difference between the curves shows what would happen to the whole market if remnant demand did not exist. Essentially, when an ad is properly targeted, it is no longer an ad, but important information.

Any targeting technique is meant to segment the content and audience at a finer level to match them together. With that, you can clear inventory and raise prices. Anything and everything that better targets helps bend the curve and is worth a lot of money.

How much would that be? For argument sake let’s say just 10% of all current remnant ads would remain premium ads with better targeting. We know in 2010 there will be about $25 billion spent on digital ads in the U.S. Approximately $15 billion of that is in search ads, $7 billion is in the display arena and the balance can be found in behavioral and other means (e.g. classifieds etc.)

We also know the average for direct sell ad placement costs about $15 per CPM whereas running through a DSP will cost you far less — in 2008 it was $0.26 per CPM and falling. As such, I will use $0.15 per CPM for this analysis.

You can use these numbers to arrive at the overall number of impressions both premium and remnant. Then take just 10% of the remnant ads and price them as premium at only half the going rate with better targeting. That’s $14 billion.

That’s right, keep just 10% of remnant ads as premium and we all make $14 billion more than we otherwise would have.

We face a choice in the digital ad world. We can work to arbitrage the match between content and ads in faster ways, using slick mathematics and Web scale infrastructure as we watch the price per ad continue to fall as a result.

Or we can find new ways to revive creativity; test and show brand lift in the digital arena, bring more brand advertisers to digital and, most important, improve the segmentation and match between the content and the ad.

Those who consider the absence of behavioral targeting as less accurate targeting are misinformed. There is so much more that can be done with analyzing content.


  1. I think the author has confused “targeting” with the law of supply and demand…an ad doesn’t magically become targeted because someone wants to pay a premium for it…


Please enter your comment!
Please enter your name here