by Scott McAndrew on March 11, 2008
Here’s two books which I definitely recommend:
Call to Action: Secret Formulas to Improve Online Results
Somehow Call to Action escaped my reading list for years. I learned about it from someone I recently interviewed. The book is based on Bryan and Jeffrey Eisenberg’s concept of Persuasion Architecture—a methodology which, at a high level, combines consumer psychology, business goals and website design with an overall goal of maximizing conversions.
Web Analytics: An Hour a Day
I’m one of those people who can’t get enough of web analytics. Written by respected blogger and Google Evangelist Avinash Kaushik, this book demystifies web analytics in an easy to read, enjoyable manner. The practice of web analytics isn’t as simple as it may appear at a glance. Avinash’s book provides new perspectives in a concise manner.
by Scott McAndrew on April 4, 2007
Google opened up its Website Optimizer (beta) to the public (anyone can sign-up). Website Optimizer is a fractional factorial experimentation tool (also referred to as a multivariate testing tool) which allows marketers to test various facets of the content and/or design on web page(s) their visitors view. In comparison to A/B testing, fractional factorial experimentation allows testing of a greater number of variables at one time while still providing reliable testing results.
The cost for Website Optimizer? It’s free. The catch? Depends on how you look at it, and more importantly, what your needs are. Things to consider:
- It appears, although nothing in the documentation I read says it explicitly, that you can only run experiments on a single page (as opposed to a flow which consists of multiple pages-like a checkout process made up of 4 screens, for example).
- I could not find any documentation that ensured me that users who were viewing a specific version of a page would see that exact same version of the page if they returned during the testing period.
- Website Optimizer does not allow variable content to be a part of the test. You can apply Website Optimizer ‘experiments’ to variable content pages, however, the variable content itself cannot be tested against.
If you’re looking to test a single landing page which is well trafficked, or a part of a marketing campaign, Website Optimizer is definitely priced right. And, there is no doubt that multivariate testing is a more effective way to perform these types of tests. If you’re looking to perform more involved testing, you’ll likely be looking to other providers such as Offermatica, Optimost, or SiteSpect. Be warned: those tools are decidedly NOT free.
