The Future Of Premium CPMs

Ben Kunz wrote an interesting article for Business Week that was published online today. The main point of the article (in my view) is that publishers accustomed to charging high CPMs on ads (sometimes over $60), are in for a rude awakening, because advertisers and ad networks can use retargeting to find the same users for a tiny fraction of the price. I disagree with Ben, but I don’t want to speak for him, so if you’re going to read this post, please read his article first.

My own experience as a media buyer and a former employee of the mother of all premium online publishing companies, CondeNet, leads me to see things a lot differently.

To be sure, there is a massive disparity between the cost of buying an ad on a site like, say, The New York Times, and buying an ad on a large network. And yes, you may hit the same users… in fact, through the magic of retargeting technology, you can sometimes be *sure* of hitting the same users (or at least, users surfing the web with the same computer and browser).

However, that’s not the whole story. It’s important to consider the following:

Branding. It’s not all about getting as many clicks as possible, at least not for all advertisers. When the publisher has credibility with the targeted demographic, then the advertiser sometimes buys media just to get across to the demographic the idea that there is a connection between the two. For example, a new stylish and expensive brand of designer women’s clothing, at the level to rival Bergdorf Goodman and Yves St. Laurent, is going to run ads on Style.com as part of their launch strategy. They’d be crazy not to, even if they already know that they can reach most of the same users on Page Six for less money. The goal isn’t always economy of scale; sometimes it’s simply brand positioning.

Placement. Not all sites are created equal. In fact, most of them are created rather poorly, especially insofar as ad placement is concerned. Do you want your ad to appear far from the content, or alongside another ad of the exact same size? It it okay with you if the site is set up to refresh the ads every thirty seconds? It’s very comforting, for both the media planner and the client, to find the ads on each site on which they are supposed to appear, in real time, and to verify that it seems realistic for the targeted users to find them. Sometimes, that level of comfort can be worth quite a bit of money.

Hands-on analytics. When you run ads on a network, you are usually not allowed access to the raw data that shows performance on a per-site basis. After all, if you find out which sites are giving you the best performance, you may attempt to circumvent the network and go directly to those sites. Instead, you get analytics based on site type, e.g. “here’s how your ad did on the men’s sites.” If you don’t have the resources, or the interest, to dig further, then results like that may work for you. But I’ve found that there’s a lot of value in knowing exactly which individual sites respond to a certain ad. Sometimes, unexpected revelations come about. Just as Ben notes that ads targeted at women are sometimes found (via network technology) to be more effective on men, I’ll note that there is no limit to the degree that you can slide and dice the data – if you have it – and still keep finding valuable info that will impress your client and save you both money on the next campaign. Will your ads do equally well on GQ and Esquire? Nerve and The Onion? Epicurious and Food Network? Are you sure? Some of my most exciting revelations and greatest successes as a media planner came from poring over the differences between two ostensibly similar sites that had yielded vastly different results for the same campaign.

Customized publisher targeting. The New York Times website offers a type of targeting called a “surround session.” For a certain period of time, all users who enter the site through certain sections (e.g. the Technology section) are exposed to the same ads, over and over, no matter where they go on the site, five pages deep. In my experience, the click-through rate, as well as the conversion rate, was astronomical compared to almost any other site, every time I bought it. Of course, you have to know which products, and which ads promoting those products, are appropriate for that context.

Customized publisher promotions. Many publishers will work with advertisers to either create customized content, or to give the ad special positioning, or even to collaborate in the creation of the ad, so that it resonates with their readers. When it works, it creates an experience that the user remembers for a long time to come. I still remember the excellent Showtime ad I saw on Nerve a couple of years ago, advertising The Tudors. The ad was a fake personals ad, as if created by Henry The Eighth. It was rendered in the style of ads on Nerve Personals, with just the right note of cynicism and whimsy to appeal to the site’s demographic. I watched the first season of the show because of it.

Advertising on large ad networks is certainly much cheaper than advertising on premium sites. And there’s a reason. Not every situation calls for premium ad space, custom content, publisher collaboration, and improving the way your brand is perceived by your target audience. But some do. And the recession doesn’t mean that advertising has to get cheaper; it means that advertising has to get better.

image by simondoggett