I sat in on a Webinar about the Google Website Optimizer and here are some thoughts and notes.
I’ll be the first to admit that I don’t like many of the pages on my website. I just don’t. Many of the designs are 3 years old, may have some misspellings, and are basically done by someone with little design experience. The problem is that the pages work. Some of the best websites in my industry look like they were done in the mid 90s. But they work, and maybe they work better than some of those shiny new sites that are built with lots of venture capital. Anyhow, there have always been ways to make changes, and determine if it worked or not, but it’s never been this easy.
Enter Google Website Optimizer. You can track many different version of your pages, and really take any of the guessing out of interpreting the results. This is a free product from Google, the only requirement is that you have a Google AdWords account. Just because you need an AdWords account doesn’t mean that you need to do it on a page that you are sending AdWords traffic to. That opens this up for any of your sites out there. Did I mention this was FREE.
Features:
A/B Testing - 2 completely different versions of a page
Multivariate Testing - For every section of a page changed, it makes a unique combination. So if you change a header, a graphic, and a description, there will be 9 difference versions of that page tested with results.
Followup Testing - Once you’ve gone with the best based on the results above, are they still working?
Testing Possibilities:
- Info rich or short and sweet
- Left or right layout
- Product facts or aspirations
- Lead or sales focus
- Reinforce site security or privacy
- Highlight a free service
In doing your tests, they recommend that you have about 100 conversions (click, email, sale, lead, etc) per version you want to test. So if you are only getting about 200 conversions in a 2 week period, you should only test 2 pages. This isn’t a rule, but I guess it makes a difference for how solid your results will be.