• by Mark Naples , October 31, 2003

I don't see how a positive answer to that can really be disputed. I've always thought that the real promise of the Web for marketers and publishers alike resided in the increasingly precise segmentation that self-selecting readers (and buyers) provide. Advances in software development make it possible to take that so much further and discern not only what someone is reading or buying, but what they were reading or buying - or considering - before.

Those of you who have been in this business think I'm a little daft now. None of this is news. But here is where the fun part is. Why is behavioral targeting okay now, when profiling across a network wasn't okay just two years ago?

And don't tell me it's because nobody is gathering personally identifiable information and merging it with the click stream data or taking it offline. This has nothing to do with PII or privacy issues. It's something far more essential than that. It's about how media is supposed to work. The only thing that really makes a difference between the profiling of 2000 and what we're now calling behavioral targeting is that one term: self-selecting.

It's the user, stupid!

Why do brands matter? Why do opt-in email campaigns so outperform opt-out email campaigns? Why does the Do-Not-Call Registry get millions of participants and why does the Do-Not-Spam Registry have politicians jump on a bandwagon in anticipation of the millions who will sign up for that? Why do Web users hate spam so much?

And why do so many marketers have trouble understanding why consumers feel the way they do?

Because users, a.k.a. readers, a.k.a. consumers, know that they deserve control of the relationships they maintain, even the commercial relationships. The interface of a masthead is no different than the front door or my phone. If a consumer chooses to walk through that door, read that masthead, make that call then that consumer will almost assuredly accept what goes along with the deal. People understand this more than many marketers would believe - which is why there was such a profound disconnect around the popularity of Do-Not-Call.

In 2000, and 2001, when I would visit Senate offices to lobby on privacy-related matters, staffers would tell me that they were more interested in the consumer affairs elements of the issue than they were the business side. I would argue that consumers go online the same way that they would visit stores or read their daily newspaper. Consumers will not frequent stores that have staffers following them down the street afterward. But, they don't mind it when attendants check on them in the stores, or when they get discounts from their local supermarket for providing their PII.

In other words, if you take care of the business side, the consumer affairs issue will all but disappear. This is because any of these quid pro quo-style relationships, if they are sustainable, are asserted and maintained by the Consumer. Consumers have to dictate these relationships on what they feel are their own terms, and they have to feel like they can opt-out at any time.

That's one reason why behavioral targeting is so interesting now, especially as companies like Tacoda practice it today. Tacoda is an especially interesting case for a number of reasons, among them is that they sort of invented this publisher-centric approach. (Remember: The relationship is between the user and the media, not the user and the advertiser.) Another key reason is that they understand that it's not just behavioral targeting but Audience Management that's creating their client's successes.

Advertisers need to ask how fresh a given audience segment is. They should know if they can target the segment in session when they are interested in their product (i.e. someone visits travel and on the next page they can get travel ads.) They should know what kind of reporting and tracking they can do on these segments, what's the reach of this segment, what are the demos, how loyal are people with this behavior, what areas can I target them in etc. For small campaigns, this might sound arcane. But, for large campaigns that are run on large, branded sites, it's easy to see how behavioral targeting can make such a difference.

To my knowledge, Tacoda has published more actual, public success with quantifiable results than others in this space. I'm not talking about leveraging better data to improve campaign performance by 50%. Some of their case studies report gains as high as 2200%! Even without ridiculous gains like these however, the evidence continues to mount. Audience-targeted campaigns in which the publisher owns that user out-perform nearly every other form of online advertising. Let's see what happens when we have cases that include rich media too.