Often, Google develops a product and then watches as users innovate new uses that are exciting in their creativity and utility. Well, we just came across an excellent example that is close to our hearts. As you may know, Google Website Optimizer was recently made accessible to all AdWords users. We've mentioned it a few times in this blog: it's a multivariate web page testing tool that can help you understand which web page designs really work. It uses a javascript code snippet very similar to Google Analytics'. For example, you may wonder, do more people buy products when your conversion link is: a.) a white button with "Buy NOW!" in lowercase, black, italicized Arial text to the right of the product name, or b.) the word "Purchase" in blue as a link with a blue underline to the left of the product name. What if you found out that version b. saw 70% more clicks than the version a. Wouldn't you use that version for all your pages? With Website Optimizer, you no longer need to guess. Instead, implement both designs and let your users tell you which works better. And then watch these small experiments affect your bottom line. However, Website Optimizer reports on only one metric -- conversions on links. What if you'd like to know how different versions of the same web page affect other metrics, such as Exit Rate, Average Time on Page, Navigational Analysis or New vs. Returning visitors. Now there's a way - you can integrate Website Optimizer with Google Analytics by making a few modifications to the javascript and to the tracking on the page. Shawn Purtell of ROI Revolution has documented a way to integrate the two that allows some very powerful analysis and action to be taken. For instance, you know that it takes roughly 45 seconds for a visitor to read a certain blog post on your site, but you've noticed in your Google Analytics reports that the average time on page is less than 10 seconds. Clearly, people are seeing your page and leaving before reading the content. Try a few different designs with Website Optimizer, and look at the results in Google Analytics to see which version works better. This is an unsupported technique, but the folks over at ROI Revolution, a Google Analytics Authorized Consultant , would be glad to walk you through it and have fully documented it here . Great job Shawn, for some innovative experimentation.Posted by Jeff Gillis, Google Analytics Team