Google Web Optimizer – Part 1

During December 2008 at TAKKLE I began to explore Google’s Website Optimizer tool. It seemed like a great tool to do testing for a lot of projects I was involved in. A lot of business decisions can come through knee jerk reactions and a lot of those decisions get passed down from “HiPPO”s, Highest Paid Person’s Opinion’. [Here is a fun article on how to train influence your HiPPO] Testing on a site is vital to prove why some of the decisions are right and some are wrong.

I began by reading “Always Be Testing: The Complete Guide to Google Website Optimizer” which was given to me by our marketing department who knows the author. While the book is a bit heavy on marketing concepts, it gives many valid reasons for testing, testing stratigies and served as a great guide to what GWO can do for you for free and what it’s limitations are.

I gave a presentation to the company (I will soon share this through Google Docs) where I used an example of a few ways we  could change the homepage to see if we could achieve more conversions to our register page.

The main concept behind the WO engine is testing either completely different versions of a page (A/B testing) or testing variations of sections on a page (Multivariate testing) to get the used to a specified conversion page. There are ways to test driving users through different funnel pages to get to that conversion, testing time on site, conversion on click and conversion on form submission. I put links to those conversion tactics since they are a bit harder to set up and are pretty buried within the WO help files

The WO can be tied to your Google Analytics account, but not have different account profiles attached to it. There are quite a few issues that need to be resolved, or other things that would make their product so much more robust, mave they will be announcing some of these in the future via their blog http://websiteoptimizer.blogspot.com/. But for free, this is a great testing tool that everyone involved in UI, testing and juggling peoples opinions should be using.

The reports are pretty simple compared to what Google has in their Analytic tools, there are 2 pages listing conversions and page section conversion with comparisons to how the different combinaiton beat or lost to the origianl. I have included my result for the experiametn I ran on an interstitial on TAKKLE.

  1. No comments yet.

  1. No trackbacks yet.

You must be logged in to post a comment.

Valid XHTML 1.1