Google’s Website Optimiser – picking the right opinion

Google’s Website Optimiser is the latest addition to Google’s suite of web improvement and traffic generation tools, and has been available for a number of months now. We’ve been using it and so far we’re very impressed. Before we can understand where this new tool fits into our web marketing strategy, we first need to understand the problem.

So what’s the problem?

It’s that whenever changes are made to a website they are typically made in an environment where there are multiple opinions about which changes are the right ones and the changes that will improve performance.

Let me explain

Imagine this scenario: you or your web agency have studied your site’s analytics data and have identified a number of key pages that simply aren’t performing and are detrimental to your site’s ability to convert your traffic into sales or enquiries. You know that if you can improve the performance of these pages you can significantly improve your site’s overall performance. After discussion with your web agency you produce a list of possible changes that you feel may improve the performance of these pages. But how can you accurately determine which of these changes have the greatest impact on performance?

Enter Google’s Website Optimiser

Google’s new tool allows you to trial multiple versions of page content while comparing the performance of the new versions against the performance of the original. By adopting this approach, you can completely eliminate the guesswork and speculation typically associated with web improvement projects.

The tool gives you two options for testing

Conventional A/B Split Testing – this variety of testing is best suited to sites with low traffic volumes. This method involves creating a completely new version of the original page and testing them side-by-side. It allows you to quickly determine which version of the page performs best.

Multivariate testing – this method involves identifying a number of elements on the original page and testing the impact of rotating and combining different versions of these individual elements. For example, you may decide that you’d like to run a test with 10 different page headings, five different images and five versions of the first paragraph. Google’s Website Optimiser will rotate all of these elements and identify which combination yields the best results.

What sort of results are we seeing?

We’ve used the tool to run a number of tests. One in particular really stands out as a demonstration of how effective this tool can be and involved a redesign of an enquiry form on one of our client’s sites. This form was then run alongside the original to determine which version performed best. The new form alone has resulted in a 40 per cent increase in the volume of enquiries that are being submitted.

What does it cost?

The tool itself is provided free of charge. The real skill lies in deciphering your site’s statistics, identifying problem areas and setting up the tests. However, when used correctly this tool can produce cost-effective and significant improvements in your site’s performance.