Almost two years ago, while all the people were in love with adwords (with SEM in general, but mainly with adwords) I was talking about something I called DAD (or desired advertising). This “theory” proposed that segmentation was key on advertising, but not just a geographic segmentation, a deeper segmentation that allows that a particular person receive the advertising impact, at the right moment, by the right media and vehicle, and if we are lucky and it really effective hi not only will allow this ad impacts him (down the psychological barriers) but also recommend it to other people that may also be looking for the same product or service.

If we can do this, the audience receive what they are looking for at the right time (it’s received as a service), the publisher receive more money for their advertising space, and the advertiser get better results from their marketing budget. Everybody is happy…

Sometimes when something looks very difficult, humans prefer to say it is impossible, so then is not necessary to spend a lot of time planning, thinking, and committing mistakes. I’m sure that is not possible to learn if you do not commit mistakes. Someone says, “Success is going from failure to failure without a lack of enthusiasm”. I do completely agree with that. You guys know exactly what I’m talking about…

What have all this to do with Analytics 2.0? Well, as a first answer (a really simplistic one) I would say, “all to do with Analytics 2.0, because if the market is changing, is necessary that analytics follows the market and develop techniques to measure it under the current circumstances”. That answer is completely valid, but I will try another one, the one I’m 100% agree with.
The valuable thing on switching to this kind of advertising is that if we are able to really segmenting using valid variables then we are going to be able to use statistical predictive models to forecast our campaigns, and if we have a learning process (databases with information from the past) then the results could become very precise.

We are going to be able to take one variable as the independent, for instance impressions (this is just an example to demonstrate, in a very simplistic way, what’s my point), and other dependent variables to forecast our campaign with extremely accurate results. You could say, I’m already doing that, and I’m sure about that, but by identifying your target audience (I mean, mapping the variables that represent their preferences) you will reduce the “black box” of human behavior. In other words, as better as you know the people, as much you know how they will behave under different circumstances (and that’s the environment that statistics love ). Like someone says… “Tell me how you measure me, and I will tell you how I will behave. Measure me in an illogical way; don’t complain when I act illogically”

I’m writing some white paper with practical examples about this, and will be available in less than a month.
The free white papers are going to be available just for users with more than a month from registered, so if you are not, do it right know by clicking here.