Follow on Twitter
Subscribe by RSS
Behavioral Targeting has a new face these days and its called Adadvisor, from TargusInfo.
In fact they prefer to call it Predictive Targeting, which makes sense because little or no click stream data is used to predict the consumers behavior. Based on historical consumer data from online and offline behavior, AdAdvisor provides advertisers the ability to target consumers like never before. Behavioral Targeting experts like to think that they have a complete and robust database of information to properly segment and target consumers, however without the additional offline data, they simply fall short.
TargusInfo has been building their databases to handle much more then simply phone number activity and connectivity, but now has a probably one of the largest databases of consumer buying history. Combining that history with segmenting and statistical based scoring they now can, with relative accuracy, determine what you are more likely to buy. Based on a segments history they can decide to show the most relevant ad. Instead of showing a free web hosting offer to you they may decide, based on your purchase history and demographic, an ad for a home remodel may be more appropriate and successful.
This is an amazing revolution in the online world. The consolidation of offline information for online use is only something that TargusInfo could pull off.
Taken from the Targus website:
No Behavioral Targeting Data Required
You need no behavioral targeting or clickstream tracking data to deliver hard-hitting, click-prompting online advertising. Where behavioral targeting fails 80% of the time due to lack of data, we let you deliver the most relevant display advertising from the moment you encounter a user.
Predictive Offline Consumer Information is Now Online
Where many services promise you an audience based on their general web behavior, we deliver an audience based on actual offline behaviors — without tracking or exposing any personally identifiable information.
AdAdvisor services categorize Internet users into anonymous population segments — in a privacy-friendly manner — based on offline consumer information and verified demographics.
Recently Google implemented its own Behavioral Targeting on Adsense called “Interest-Based Targeting”. As if you didn’t know you are being watched and your behavior is being documented and segmented. Do you have an issue with this? I am open to these types of developments because I will no longer see ads that I am not interested in. I guess the deeper question is can Behavioral, Interest-based or predictive targeting begin to form your buying habits for you?
Many times consumers simply need a little nudge to begin a buying habit. If that is all that is needed to push a person down a specific habit path I believe I could ultimately, with in reason, force someone to begin buying specific products. Just a thought, however if Targus can determine that consumers that are purchased a coffee maker are more likely to buy a new china set I simply would need to get a person to buy a coffee maker and then show them an ad for a new china set. Now, I over simplified the process, however you can determine a path that leads to your product you would then simply need to send them down that path. That could take a days, months or years, but with certain amount of accuracy you can direct people to specific products. And if that is the case, at what point does the information get skewed? When would a targeting bias affect the relevance of the consumers habits and would it really matter if it did anyway?
.
.
.
3 Comments
Dubious value. If they’re using an IP lookup table and clustering people according to zip-code, it won’t tell them much except alot of people work in downtown areas and alot of servers are based in El Segundo and Burbank. To extrapolate and say that the person with an IP address that resolves to El Segundo tends to behave like people who live in El Segundo, well good luck with that during the workday. The person might live in Downey or Beverly Hills and work in El Segundo. Or, they may live in Reseda, work in Culver City but their server farm is in El Segundo.
Now, if this is more than resolving IP addresses, how are they doing that without clickstream data from cookies?
The vaporware announcement reminds me of LeadID. How’s that coming along?
[Reply]
JoJo,
Take a look at the link in the post and possibly watch the video. They are working with far more data then simple IP address locations. Offline buying data and habits are being used.
This ads more data to the mix. Example, I buy more products offline then online. The information is highly targeted because if it.
Maybe Targus can confirm.
[Reply]
JoeJoe –
The product is indeed live and being used by 5 of the top ten advertising networks in the US. We previewed the solution at the Online Lead Quality Summit last fall, and formally launched it at OMMA Behavioral in February of this year after a very successful private beta.
As you correctly stated, doing anything by just using IP address is of dubious value. There are already several companies that build IP lookup tables to city-level location data, two of which are Digital Envoy and Quova.
Secondly, the problem with using IP address is that it isn’t static enough compared to cookies. There also continues to be a concern about privacy and IP addresses. The EU considers them PII.
In the AdAdvisor solution, we have built a targeting methodology that delivers a predictive score across hundreds of behaviors and lifestyles, as well as individual-level demographics in each cookie.
Our coverage solution uses trusted partners that give us access to the browser during a registration event / login / or email open, during which we lookup the predictive targeting information and demographics and place the anonymous information on the users browser in a cookie.
We do not collect clickstream data in the cookies, and have built a solution that is more predictive and generates more lift than the commonly deployed optimization solutions.
All of our methods are compliant with the Network Advertising Initiative guidelines.
Happy to discuss further if you have any questions past your general expression of doubt.
[Reply]