Getting good behavior from your ad network

Media planners dwell in a quandary: the efficiencies presented by ad networks and the effectiveness promised by behavioral targeting make perfect sense, but there are hundreds of players in this space and differentiating between them can be difficult.  How good is their inventory?  How good is their targeting?  How can you know — in advance or even after a campaign — whether you’ve made the best choice?

Media planners also inherit some of the challenges faced by the networks and targeters themselves.  Is a publisher giving good inventory and suitable context?  Is it treating all partners fairly?  How do you verify?

This tangle gets manifested in the "˜guts’ of many web pages.  For a review of how pages are put together and the many URLs that build into a single page, see A look under the hood of Ad Impact. The advertising infrastructure of the page below is more tangled than most — as the page loaded, it spawned calls to multiple ad-network-related domains, including: quantserve.com (quantcast), revsci.com (revenue science), yieldmanager.com, advertising.com (an ad-serving division of AOL), turn.com, lucidmedia.com, dotomi.com, admeld.com, and pubmatic.com.

The result in this case illustrates media planners’ fears: odd and botched placements. An AirWick horizontal banner ad got served into a vertical "˜skyscraper’ spot.  Figuring out who’s responsible when so many parties were (or could have been) involved would mean teasing out the tangle.  Certainly, the site owner bears some responsibility for the ad operations snafu, but with so many actors involved, it’s really hard to determine accountability, and occasional mix-ups seem almost inevitable.

No matter whose fault it was, the advertiser is the obvious victim.  So how can an advertiser keep such things from happening — or prevent less obvious snafus like "˜targeted’ ads served to the wrong target — without having to abandon networks and the advantages they offer?

We spotted a pair of best practices in play side by side earlier this week while studying ad networks and the daisy chains that serve up so much of the Web’s ad inventory (and often quite effectively, we should acknowledge).  Looking at the two ads on the Snopes.com homepage below, one might well wonder whether they’ve been served as the advertisers would wish.  The innards of this page raise concerns about possible tangles — several ad networks and network optimizers have a hand in the serving of these ads, including contextweb (and its ADSDAQ ad exchange), fastclick, burst, bluekai, and casale media.  So are the Sun Life Financial and Verizon ads targeted?  Or misplaced?

By coincidence, these two advertisers are using two very practical methods to address the uncertainties that come with using networks as a campaign strategy.  One is utilizing an ad verification service, which will report the pages where the ad has been served (a service some networks provide on their own, but most don’t) and where the ad appeared on those pages.  Meanwhile, the other advertiser is enabling ad effectiveness surveys; they will be served to a portion of the audience to measure whether the network-run ad raised awareness of the brand and good feelings toward it.

Using ad networks entails a loss of control and the risk of unknown outcomes.  But relinquishing control is the only way to take advantage of the networks’ many options for reaching a broad audience cheaply or for honing in on a specific consumer segment across publishers.  We don’t see a way to escape the media planner’s quandary — in fact, the increasingly complicated relationship between publishers, networks, and consumers only deepens the quandary.  Services like ad verification and ad effectiveness provide an important way to maintain control and to ensure that using a specific network yields the desired result, and not a tangled mess.