Admeta Answers Questions About AdImprover
Admeta answers questions about their advertising optimization solution.
Could you tell us more about the methodology behind AdImprover? How do you apply genetic programming to marketing?
The method is based on the same mechanisms you find in nature, i.e., evolution. A simple explanation is that a computer program is trained in selecting the right creative for the right visitor. The worst-performing solutions are removed from the decision-making process, and the best ones are combined to produce new alternatives to fill up the void left from the removed ones. That way the “genes” from the good ones are left, and the process mimics nature. This means that the system is self-learning and quickly adapts to changes in user behavior.
Why would your methodology perform better than choice modeling, optimal design, or the Taguchi Method modified for marketing?
The major difference is that we don’t treat all visitors the same. AdImprover makes a unique decision for each individual exposure and learns from it. AdImprover is adaptive and changes its behavior when the user’s behavior changes. The way the visitor is acting can be different at the end of a campaign compared to the beginning of it. Another important difference is that AdImprover doesn’t need that much data before it kicks into effect. Between 200 and 300 conversions is enough for the turbo to kick in.
Would you do any head-to-head tests with competing methodologies?
Absolutely! We would love to. We have done our own tests against statistical algorithms, and they didn’t stand a chance.
Do you find it difficult to convince prospects that they should apply science to what to many has been considered merely a creative endeavor?
Yes, unfortunately. The problem is that it sounds too good to be true. We have, however, solved that in a simple manner by changing our pricing model to a pure revenue-share based on the improvement. That way the customer isn’t taking any risk at all. We have never shown a lesser improvement than 30%.
When using AdImprover, how much of it would you say is technology (including methodology) and how much is people?
Up until the campaign is launched, there is some human involvement. For instance, you have to produce different creatives and you have to define what AdImprover should optimize based on (orders, revenue, clicks, etc.) Once the campaign is launched, it is 100% technology. The system optimizes better than any human could. It just gets better after each exposure.
Testing and targeting:
Optimost, TaguchiNow, Offermatica, Memetrics, SiteSpect, wunderLOOP, Conversion Multiplier, Maxymiser, Google & TouchClarity
![[del.icio.us]](http://www.webanalysts.info/webanalytics/wp-content/plugins/bookmarkify/delicious.png)
![[Digg]](http://www.webanalysts.info/webanalytics/wp-content/plugins/bookmarkify/digg.png)
![[Facebook]](http://www.webanalysts.info/webanalytics/wp-content/plugins/bookmarkify/facebook.png)
![[Google]](http://www.webanalysts.info/webanalytics/wp-content/plugins/bookmarkify/google.png)
![[LinkedIn]](http://www.webanalysts.info/webanalytics/wp-content/plugins/bookmarkify/linkedin.png)
![[MySpace]](http://www.webanalysts.info/webanalytics/wp-content/plugins/bookmarkify/myspace.png)
![[Reddit]](http://www.webanalysts.info/webanalytics/wp-content/plugins/bookmarkify/reddit.png)
![[Slashdot]](http://www.webanalysts.info/webanalytics/wp-content/plugins/bookmarkify/slashdot.png)
![[Sphinn]](http://www.webanalysts.info/webanalytics/wp-content/plugins/bookmarkify/sphinn.png)
![[StumbleUpon]](http://www.webanalysts.info/webanalytics/wp-content/plugins/bookmarkify/stumbleupon.png)
![[Technorati]](http://www.webanalysts.info/webanalytics/wp-content/plugins/bookmarkify/technorati.png)
![[Twitter]](http://www.webanalysts.info/webanalytics/wp-content/plugins/bookmarkify/twitter.png)
Posted in Test and Target, Web Analytics | Comments Off
Subscribe to RSS feed
Subscribe by email
