Data Freedom in the New World

American’s are a very eclectic group of people. I don’t think anyone would dispute that. I also don’t believe anyone would argue with the statement that within societies, human beings tend to have certain things in common; People travel, drive cars, pursue some level of education, etc (widespread trends). While we celebrate the diversity of people in our society, I find it very interesting that we often focus primarily on only a select few dimensions of their behaviors for marketing purposes. For example, the cross section of users who visit Kayak.com to book travel or visit the NYTimes Travel Section represents an audience with a very wide range of individual interests. There is certainly a value in being able to identify a “travel user” in this particular example, but my point is that many people travel. The concept of a “travel user” in this context is more a representation of a specific “in market user” than someone with say a particular interest in geography or travel blogs which I believe are two distinct things that are too often lumped together. The idea of “in-market” data has quite a bit of merit if used properly, but in the process of media buying it is often leveraged in a way that is counterproductive to the economic goals of the advertiser. Does the fact that someone visited Edmund’s because they are looking for a car to potentially buy makes them an “auto user” in a general sense of the word? No, it just means that of the 50% of American’s who have a car, they happen to be in a small sub segment of that group considering getting a new one. These people should not necessarily be associated with “auto enthusiasts” (ie people who visit Wrecked Exotics or Car and Driver, etc). To compound the issue, many of the major data companies aggregate users from their various publisher sources and compile “meta data” around a particular segment they arbitrarily define (I am magnifying the lack of classification segmentation a bit for the sake of proving a point). This is paradoxical to the whole purpose of a data market. The use of data is supposed to take the intuitive human component away from targeting, but I don’t believe that is what is really happening. Behind the glossy marketing collateral and sexy flow diagrams, a majority of the data market is really a collection of business development guys striking deals with publishers to place cookies on their users and then finding a way to aggregate that in an intuitive way. They operative word here is “intuitive”. Intuition is a human trait that you won’t find noted in any description of scientific method.
The online display market has been plagued by a constant battle between scalability and granularity. However, the growth of supply and demand aggregation has alleviated some of the operational hurdles to achieving both at once. That being the case, why is data being approached using a methodology that media buying has already abandoned (To be clear, I am talking specifically about the bundling of data). On the media side, advertiser’s now have the ability to access huge amounts of differentiated publishers and content through dramatically less platforms (and emails, etc). This empowers them to abandon the traditional network model and not be forced to bundle their campaigns in with countless other advertisers using a technology that must appease publishers as well (since they are also clients of the same intermediary). My question is why can’t data providers do the same thing? The existing platform infrastructure offers the scale required for data providers to act individually, yet the trend towards the network model persists.
Arbitrage only acts to reduce value for the end participants. Combine this with the notion that people are extremely multidimensional and we can clearly see how there is an opportunity for large niche publishers to take ownership of their data and not allow it to be watered down and bundled. Let’s use the site VampireFreaks.com as an example. You may laugh, but if you consider the various dimensions of a person, the ability to identify a user from this type of perspective represents a very different type of information than the concept of an “in-market travel user” does. This view of information should be very valuable to a marketer as it is roughly the digital analogue to the way product placement is used in television shows or magazines catering to a particular group of fans or enthusiasts. In many ways, it could be argued that this data might have more long term value with regard to developing consumer brand loyalty than targeting audiences of “in market users” in relatively widespread behavioral groups. Glenn Beck, Jamie Lynn Spears and I might all book plane tickets on occasion, but beyond that I don’t think we have much in common.
There is a huge opportunity for large, niche publishers to take ownership of their data and empower a new dimension of the data marketplace. One that is both lucrative to them and differentiated to the media buying community. The infrastructure is in place, so the only excuse for archaic methodology is a lack of education.