Google Website Optimiser

I've been fiddling around on a couple of sites with Google's Website Optimiser (or you might get "did you mean Google's website Optimizer") and thought I'd share a few findings.

First of all, I think it's a very, very valuable tool which can save you hundreds of Euros in testing and public opinion polls, etc. The idea is that you can create very similar pages to your website and then test them. You need to add a bit of javascript code to the test pages you want to run the tests on, and the website optimiser starts to compile a report.

There are 2 types of test you can do. A Simple A/B split test and Multi-variate testing which basically just means you test various different things on the page at the same time. The great thing about website optimiser is that you can test absolutely anything.

I'll give you an example of how an A/B split test can work. If you choose the goal page as your conversion page and make some drastic changes to your landing page from a Pay Per click campaign for example, you can track not only which ads are working for you, but which version of your home page is creating more conversions, too. It doesn't always have to be your home page either - you can literally fiddle around with any page on your website - even test run a completely new design.

Multi-variate tests are much more specific, and generally take a lot longer to give a good result. As with most experiments, it's important to be able to interpret the results. Google makes this very easy with a percentage of the possibility of beating the original test page. With the multi-variate testing, the interpretation of results is slightly more difficult but still great to see.

A small criticism which I've found. There must be something in the javascript that blocks the referrer stats in some of the website statistic programmes. I have a query open in Google groups and will post a fix if I get one!

For more details about SEO in Barcelona, or in-house SEO training in Barcelona please contact me.