The threat from targeting
If one promised “revolution” in advertising online has been the ability to pay only for ads that perform, the other has been the ability to aim marketing messages at certain individuals with a precision not possible in the offline world. In fact three dominant models of targeting exist on the Internet, two of which have direct analogues in traditional media:
Demographic targeting uses characteristics such as age, gender, income, and education to define a desirable audience. Most traditional advertising is targeted by demographic (and occasionally “psychographic”) profile at least in the selection of venues, and the same logic applies in well‐known destinations online. If BMW advertises on the Wall Street Journal’s site, at least one reason is its audience of high‐income, educated professionals.
Contextual targeting based on the editorial focus of the venue also routinely figures in media planning both online and off. If BMW buys impressions on Kelley Blue Book or Edmunds.com, it hopes to capture potential car buyers already in the decisionmaking process.
Behavioral targeting has no traditional precursor; it relies on cookies to sort Internet users based on what sites they’ve visited, what search terms they’ve used and what actions they’ve taken. If BMW wants to target people already in the purchasing “funnel” it can buy profiles of car‐seeking consumers from a “profile exchange” such as BlueKai. Such a profile might be based on visits to an automotive comparison site, on a search for “Mercedes dealers in New York,” or on preapplying for a car loan, for instance.
Some high‐traffic publishers are able to deploy sophisticated targeting technologies in‐house. Dow Jones uses all three approaches described above to serve ads across its network of business‐and‐finance sites. Two Wall Street Journal subscribers visiting wsj.com — say, a Cleveland‐based small business owner and a Manhattan financier — may see different ads based on anonymous profiles derived from their registration data. (The profiles are encoded in cookies attached to their browsers.) Or, someone who frequently visits the “technology” pages of wsj.com may be assigned to a profile that draws tech‐related ads, even as the same user travels to marketwatch.com or barrons.com. However, most behavioral targeting occurs through an ad network like AudienceScience or a profile exchange such as BlueKai. For behavioral targeting to be predictive, it generally requires wider inputs than a single site can provide; the goal is to assign users to particular profiles (i.e., “golf fanatic”) based on their behavior across a wide swath of the Internet, and then to be able to reach those users wherever they go. BlueKai calls this “audience portability,” and makes no bones about the disintermediating effect on publishers, or “inventory”: “Today, media buying is constrained to only buying data that is tied to a particular inventory. …At BlueKai, we understand thatʹs not the way things should be. We know that online data should be separated from the media — and made customizable and accessible to everyone, at scale, anywhere on the Internet.”17behavioral targeting represents a distinct threat to publishers: By discriminating among users individually, behavioral targeting diminishes the importance of a site’s overall brand and audience profile. Suddenly the decisive information resides not with the publisher but in the databases of intermediaries such as advertising networks or profile brokers. A similar threat may be emerging in the domain of demographic targeting. As it becomes more possible to attach permanent demographic profiles to individual users as they travel the Web, the selection of outlets will matter less in running a campaign. This is why online media outlets tend not to participate in third‐party ad networks if they can avoid it. “We donʹt want to be in a situation where someone can say, ‘I can get you somebody who reads the Wall Street Journal while theyʹre on another site that costs half as much,’” explains Kate Downey. In fact, Dow Jones is looking at ways to not just resist the trend, but reverse it, by pulling outside behavioral or demographic profiles into its own ad servers. This would let the Journal and its sister sites target ads based on what their visitors — even unregistered ones — have done elsewhere on the Internet. “The goal is to be able to give a better ad experience and user experience to even the anonymous people coming to our site,” Downer says. “That would be fantastic.” Publishers may also try to control contextual targeting by packaging their inventory as “content channels” designed to respond to advertiser priorities. For instance, when automotive inventory was at a premium, many sites rushed to create new sections to draw advertising from car manufacturers, in some cases pushing beyond their normal editorial range; a good example is Forbes’ Luxury Car channel online.