For years, we’ve all heard about behavioral targeting for the web and how its going to work... but never quite does... Well, some of the techniques being used in this kind of targeting for web inventory can also be used for email – and they work quite a bit better.
How it works
Basically we need to look to track actions by users and manage those users differently.
In the case of email, the most common action to track is clicks. So if we mail to a large list and receive a certain segment of the list clicking on the creative, those users have indicated a certain level of interest in that particular topic. Some people do this with opens, but via clicks is a far more accurate indicator.
Once you have smaller targeted lists that have responded via clicks, you can focus on sending targeted offers to those segments only, or broker that data separately.
Unfortunately, most ESP’s don’t support tracking clicks by category. Also unfortunate, if you don’t track this, you’re losing a whole level of valuable data. Lets say you’re mailing to an email list, but you also have full postal data on your users. If you are tracking category clicks as you mail to your list, suddenly that postal data has a whole new level of information about it, which can be rebrokered offline at far higher CPM’s. I’ve spoken to a couple of direct mail brokers and they love this kind of data, since users have clearly indicated their preferences.
Who is doing it?
Larry Organ is a real pioneer in this space, with his company ConsumerBase. I first read about Larry in Forbes magazine (http://members.forbes.com/forbes/2006/0327/052.html). They had a rather scary article about how he was invading everyone’s privacy. While this kind of data definitely has some serious privacy implications, its not going to be as bad as Forbes made out.
Basically, Larry likes to add value to data. Therefore he takes existing data, and adds a behavioral level on top of it, by tracking clicks. He’s systematized this so well that he’s even filed for a patent on this – which if he gets it, I think will be very valuable one day.
He’s been doing this for quite a long time, segmenting his data by different interest categories. You can even buy data from him very easily through his site http://responderinfo.com. Try his data out and see if it works for you – and let me know if it works well. They have 20 million names all broken down by interest categories with full email and full postal.
Another company that is starting to apply behavioral analysis to email is Q Interactive. They recently released an email service which does behavioral email follow-up. Over time they build profiles on what users like, depending on a number of different variables. According to their press release, they use "more than 1600 unique segments, including self-reported demographic, geographic, behavioral and transactional data and category interests." So over time, Q Interactive is learning what users like, builds profiles for them, and therefore does a better job marketing to them. You can read more about their product here: http://www.qinteractive.com/news/releaseDetails.asp?ID=227
Behavioral email marketing can open up a whole host of ethical issues, as raised by Larry Organ’s Forbes article. But over time these will be worked out. And so, at some point we can expect Google to get into behavioral marketing – however it’s not going to be so easy for them. They are extremely concerned about the affect privacy issues will have on their brand. Therefore the door is currently open for many companies to enter the behavioral space. Long term, we can expect Google to become involved since behavioral targeting is likely to become the way Google can compete with traditional TV – the ads for video will only really become effective when behavioral data is added, improving the targeting.
TV Behavioural targeting
How would TV behavioural targeting work in practice? Well, you might be mailing to your list, tracking clicks by category, building up profiles of your users. You then become part of a behavioral targeting network, feeding this data into the network. As TV shows are being broadcast online, users will be shown ads during the shows. The targeting for the ads will be done based on your email click data. You’ll get paid extra for your data, and the TV shows will able to earn higher ECPM’s because the audiences being targeted will be far more relevant. Today’s version of this will be for feeding data into the banner networks.
If you’re interested in learning more about how to do this kind of targeting to your list, a great book about it is called "Drilling Down", I’ve written a review about it here:
http://www.adrianbye.com/favourite-books/
You’re also welcome to join my list at http://tastips.com, where I talk more about this topic.
So remember, track your clicks! They’re worth a lot!
Digg 
