Should Targeting Cost More?

What kinds of things do you like to do with your time?

What types of news are you most interested in?

How much do you spend online, and on what products?

Is allergy information more important to you than beach conditions when you check the weather?

These are the kinds of questions advertisers have been asking for a long, long time. Since the early days of the 20th century and the marriage of psychology and advertising, companies have been looking for ways to persuade the persuadable, and that has relied on a combination of trying to place messages in context relevant to the product or service being promoted at a time and in a way that find the potential consumer in the best state of mind for being spoken to about it.

But what we know about how or why people use media has never been very precise, so we've had to go by other indicators of what people think or like or do.

As it stands today, demographics serve as surrogates for more meaningful information about a particular audience. What behavioral targeting promises is a way to identify an audience based on not simply their statistical data points, but on their likes and dislikes as deduced from their actions.

Certainly, products endorsed by the American Association of Retired Persons can be targeted based on a 50-plus demographic, but what about something like, say, snowboards? There are kids 10 years old who want them and guys in their thirties who want them. These are demographics so divergent, I can't really target based on demographics. But perhaps these consumers share a similar state of mind: new, exciting, daring, on edge. Whether you are six or 60, if you are interested in my product, then I want to talk to you about it.

So "behavioral targeting" has become the mantra as of late. It's getting more play at conferences, in newsletters, in the trades. The upcoming Interactive Advertising World in September has a line item on its agenda specifically dedicated to the subject. iMediaConnection has a section of content dedicated exclusively to issues involving behavioral targeting and marketing.

Given the amount of time behavioral targeting and marketing has been available to the marketplace, it is now time to ask the question: is audience segmentation based on behavior meaningful?

At a conference last year, attendees in a survey indicated that it was contextual advertising and behavioral marketing that were among the most promising advancements for advertisers in the online space. That these two methods of targeting are often linked is no accident. Involvement of a user in a particular context can be indicative of a particular kind of behavior. Being able to deliver advertising based on context and behavior is and has been the promise of online advertising since the beginning. It comes as no surprise that improvements in this arena are held out as examples of being among the most promising trends for advertisers.

Behavioral-based targeting, as an idea, is not really something that is all that new. Some would say that it was the same as psychographic targeting. The fact of the matter is, however, that it has not been possible to buy media this way and actually assign an advertising message to a particular individual (or the media device they use) based on this kind of information.

Now, companies are providing technologies that are either developed organically, lifted from the leftover pieces of other technologies, or use other wholly formed technology that can be put behind the curtain, that make it possible for publishers to provide behavioral targeting to advertisers.

Real-world numbers show it works

Is it working, though? Are publishers or advertisers yielding anything from this form of segmentation?

According to some of the recent press, they are.

Douglas W. McCormick, iVillage's chairman and CEO was quoted this past May saying, "When we first made this technology available, we all intuitively thought it would work to make advertising schedules more impactful. But now it's a fact and this study proves it."

The study he is referring to is the one done by Dynamic Logic in conjunction with Deutsch for Snapple-a-Day, the recently launched meal replacement beverage from Snapple Beverage.

The technology used to segment the iVillage audience was Tacoda Systems' Audience Management System, which sifts through vast amounts of user behavior data and essentially "makes decisions" about what kinds of people make up that audience in order to best determine what kind of advertising to serve them.

Another provider of capabilities similar to Tacoda's is Revenue Science, and they, too were part of a study that was released just this past May featuring the benefits of using behaviorally driven audience segmentation.

The study was done through TM Advertising, the erstwhile Temerlin McClain, for its client, American Airlines. The objective of the campaign was to get in front of business travelers and remind them basically how sweet it is to fly with an airline that feeds you and doesn't have socialist seating arrangements, a la JetBlue.

According to the study, Revenue Science put behaviorally targeted ads on in front of 115 percent more business travelers who travel at least once a year versus those ads run on the Web against general run-of-site inventory. More frequent travelers were reached 145 percent more.

With the technology in place, Revenue Science identified readers who were visiting travel sections of the site like "Desktop Traveler" or "Takeoffs and Landings" as belonging to a travel-seeker segment and then served American Airlines ads regardless of the editorial context.

There are other media properties out there that are capable of offering something similar to behavioral audience segmentation. Yahoo! Consumer Direct, for instance, lets advertisers target audiences identified by cross-tabbing not only surfing behavior but also data from AC Nielsen that reveals a user's purchase pattern. has the remains of Personify as the engine driving its behavioral targeting efforts.

And lest we forget desktop applications like WhenU or Claria that can use both context and content engagement behaviors to inform the patterns they see and result in ads served to particular audience segments.

So, although the evidence in favor of behavioral targeting and marketing has not made the transition from hill to mountain, it is building and it does indicate that behavior is a far superior indicator of belonging to a particular audience than one's membership to a demographic segment.

The only question remaining to be answered is: are publishers successfully yielding a premium for inventory behaviorally segmented that would otherwise ordinarily be blind ROS inventory?

Seems to be that some publishers are getting incrementally higher CPMs for behavioral inventory, as far as I can tell from some of the media I've been buying. But media buyers are notoriously suspicious of paying premiums for something they cannot see (if I'm buying behaviorally segmented inventory targeted to women 25 -54 who like crafts, I'm not likely to ever see my ad).

What advertisers really want

It is behavior that advertisers really seek to address, not an age or gender. It seems stupid that what we've been doing in advertising for decades is to take a state of mind or a type of behavior, translate that into demographics, and then use those demographics to translate back into a means of reaching people in a particular state of mind or with a specific type of behavior.

Given the new technologies emerging for use in conducting online advertising, marketers won't have to any more.