• by Phil Leggiere , November 28, 2007

Truly personalized, one-to-one marketing is the ultimate destination, the holy grail, toward which most advertisers aspire. Making progress in the long journey to get there involves moving from targeting by the lowest common denominator to greater and greater particularity and relevancy. While behavioral targeting has rightly been touted as a critical leap forward in that quest, up till now the conventional thinking and practice of BT, , has remained limited by a  lowest common denominator approach to segmentation, as Andrew Westmoreland, CEO of AdRevolution, explains below.

BI: What was the motivation for developing NIME? How does it expand conventional understandings of demographic, contextual and behavioral targeting and how they can mutually reinforce one another?

Westmoreland: I asked more than four years ago: ‘Why am I constantly bombarded with irrelevant advertising?  Wasn't the Internet supposed to change all that?'  I'm just like everyone else in that most of the ads I see in my day are not helpful to my life.  I knew there was potential for a solution that produced better results for everyone, especially the consumer.

We said from the beginning that our behavioral targeting needed to be able to act on the most granular level of information.  Rather than just tagging someone as part of the ‘Auto' bucket because they are shopping for a new car, we are able to know exactly what type of vehicle they want to purchase....  Our technology constantly relearns where and how messages are most effectively being deployed.  Our ability to use a person's unique information to customize ads on the fly has already greatly increased revenues for all of our publishers.

BI: What have been the biggest obstacles to the optimization of targeting?

Westmoreland: There is a knowledge gap between the technology folks and advertisers. The advertising industry has accepted that market segmentation into 30 or so silos is enough. That's never felt right to me.  As a technologist and person genuinely interested in human behavior, I knew that a deeper method to match behaviors with ads would yield superior results.  The real opportunity here is for technology to enable individual dialogues.

I also believe that the lack of standards (especially in terminology) is a major obstacle.  I bet if you ran an open survey to the readers of this column asking ‘What is behavioral targeting,' you would get a unique answer from each respondent.

BI: What in your experience are the most misunderstood things about behavioral targeting by publishers and advertisers just beginning to develop behavioral targeting features?

 Westmoreland: Since the Web came into being, there has not been true behavioral targeting. Instead, a bunch of companies deliver crude market segmentation, which has been common practice for many years, especially in direct mail. But, segmenting 250 million people into 30 buckets is a very unexacting process. We, in contrast, self-learn on the fly based on an infinite number of behavioral variables and descriptors.

What this means is truly transformational.  It means an end to wasted ad spending. For example, instead of all of America being shown an ad for the new Ford F-150, only those who really want a new truck would receive that ad.  People interested in saving the earth or just some gas money would see an ad for the new Prius.

BI: Can you give some examples of how your behavioral platform is being successfully deployed?

Westmoreland: Most of our early publishers are lead-generation players.  For them, at the end of the day, earnings are king.  All of them have seen a significant increase in revenue. We integrate their data and assimilate it into NIME for usage in advanced micro-targeting. We have leveraged this data to lift effective CPMs at least 3x.  One of our best partners is fully integrated, and has seen effective CPMs increase from the $1 range to the $30 range.

We've also given consumers the ability to give us direct feedback about individual ads.  We receive "thumbs up" and "thumbs down" inputs and use them to adapt the advertising experience on the fly.  I have not seen any of our competition actually engage the consumer in this way.  It's proof that we really care about what the consumer has to say, and that we can actually use direct feedback information in real time.

BI: What kinds of approaches to cross-channel behavioral targeting are you developing, and how would you describe the progress of your mobile initiatives in this area?

Westmoreland: The real value in the multichannel approach is that the advertiser can reach the consumer where he or she is most likely to respond.  A great example is that a consumer may positively respond to a banner ad, but not fully convert.  A simple email reminder a day later may be appropriate to allow the consumer the chance to finish registration when it is most convenient.

It was always a fundamental precept of the software development that it would have to integrate seamlessly with legacy systems and be simple to adapt to any platform, in any channel. Currently, we are actively integrating display and email into the NIME, and one of our clients is working very successfully with mobile. Today, our mobile distribution partner has seen consistent 20% conversion rates.  We are actively negotiating with a number of advertisers to join our mobile initiative.  

BI: What are the most important challenges to effectively deploying and growing the platform you see on the horizon for the rest of 2007 and into 2008?

Westmoreland: We have an interesting and unusual challenge.  Our challenge is to fully maximize our first-mover advantage, and scale the marketplace to its full potential.