I hope Google teaches people how to properly prune experiments. The problem with taking out experiments that are underperforming is that it cuts out the information about the performance of all the individual elements on the page. Meaning some elements will have more conversion data tied to them than others, creating an unbalance of data.
Also you can seriously impact the validity of your test if you prune too early. While it's a good idea and the impact is smaller since GWO uses full factorial, Google needs to be careful and let people know about the dangers of that shiny new feature.
In response to Dallas SEO, I haven't read anywhere that Google is allowed to use the information tracked by Google Website Optimizer. I have a hard time believing that they would without consent either. If you have any hard information on this, I'd love to learn about it though, especially since my company provides testing services with our own platform.
iEntry 10th Anniversary
RSS
Newsletter
Advertising





















Too Much Information
Personally, I am worried about giving G any more information than they already get. Call me paranoid, but the amount of data they find ways to make us "want" to give them is staggering.