Previously, we discussed how over-targeting can potentially cause a loss for the advertisers in terms of lost conversions and missed learning opportunities. We explored how Grouping and Referring are the first two things we should consider before over-targeting without sufficient data and testing. This week, we will look at the third consideration: Taxonomy

Taxonomy – Do you remember the first time anyone explained Behavioral Targeting? Didn't it sound a bit like magic? "We are going to magically put the ad in front of the right person. How do we know who is right? By the exact science of closely examining their past click history." Really? Even without considering the effect of multiple users who share the same device (shared email accounts are a whole other story), let's look at how the user's past click history translates into behaviors that we can leverage. A user goes to a certain page and sees specific words on it. Taxonomy describes the words that a user will see and we associate certain behaviors based on these words. Taxonomy is a word most of us barely remember from high school or college botany courses. (This is another great example of how online marketing has absorbed the best words from other disciplines!) Taxonomy selection lends itself to the kind of wiggle room that would concern most advertisers who are guilty of over-targeting if they really took the time to learn how differently each BT (Behavioral Targeting) provider creates behavioral profiles.

Behavioral Targeting follows a principle of frequency of interaction over time. If a user visits pre-selected content (a word, a phrase, a website page) enough times in a given period of time, we assign a behavior to the user. If the user clears their cache of cookies, we can't use Behavioral Targeting. Happily, most users don't. So, for example, if a user goes to 10 sites that are women-focused about 100 times in 2 weeks, BT would guess that the user is female. The frequency, the degree of interaction and the time period are determined individually by each site, network, or 3rd-party serving the BT ads. The pre-selected content is different for each player as well. In other words, they are all using a different taxonomy and ad-serving strategy in addition to different ad servers with different built-in rules for serving BT ads. Given that, how valuable is it for the advertiser to be too specific about targeting parameters without testing a variety of behaviors? Consider that each BT provider works their magic differently, which inevitably results in traffic being served by their ads that are converting at a different rate.

Given the different taxonomy/ad serving methodologies out there, are we really so sure we want to hone in on a specific target? Why not go wider and conduct testing first? Maybe we are missing an undefended niche or sweet spot so large we could drive a truck through it.

This is the 3rd installment in a series about the dangers of over-targeting. Next time, we will look at the challenge of scaling when we start with an extremely tight targeting requirement.

The opinions and points of view expressed in this commentary are exclusively the views of the author and do not necessarily represent the views of MediaBizBloggers.com management or associated bloggers. MediaBizBloggers is an open thought leadership platform and readers may share their comments and opinions in response to all commentaries.