In my last article, I discussed a list of elements you can test on your website and landing page. One way to easily conduct tests on your website is through Google Website Optimizer. Today, I’ll give a description of Website Optimizer, how it can help your testing process, and some tips on how to use this tool.
Google Website Optimizer allows you to experiment with the content on your website in order to determine what appeals best to your audience and generates the highest conversion rate. With this tool you can test almost any element of your website including headline, body copy, images, call-to-action buttons, benefits lists, layout, and the list goes on and on. As we discussed in my last article, you should already have a list of attributes you would like to test on your site.
With Website Optimizer you can launch multi-variant tests relatively quickly. And here are a few tips to consider when launching your experiment in Website Optimizer.
You will need a Google Analytics account. When conducting an experiment, you do not need to access Analytics directly but before you can launch an experiment you will need to have an Analytics account as Optimizer uses some tracking features from Analytics.
Choose the test that is right for you. Within Website Optimizer you can choose from two different kinds of tests: A/B tests and multivariate tests. A/B tests compare two entirely different versions of your page so you can try varying layouts and move things around on the page. Multivariate tests are a bit more flexible. These types of experiments allow you to test smaller content variations on a given page. For example, you could try out two different headlines or two different benefits lists or three different images.
Choose the elements that will have the most impact. You want to test elements on your page that have the potential to influence your audience. That’s the whole point of split testing! You’re looking to generate a higher response rate. So, test only the elements on your site that will generate a response because it is this response that you’ll be using to determine the effectiveness of each variation.
- Add control script, tracking script and identify your page sections.
- Add a snippet of conversion script to your conversion code.
- Publish and validate your test pages.
I am not going into extensive details on each step because you will receive plenty of instruction when creating your experiment, but this is what you (or web development team) can expect.
Collect enough data to make a statistically valid decision. This tip is the easiest because Website Optimizer will monitor and gauge your results and let you know when you have a significant winner. As you are conducting an experiment Website Optimizer provides enough data for you to make a well-informed decision. Here is a quick screen shot of a testing screen in Website Optimizer:
In this experiment, I am testing an element on my page that includes a benefit list (against one version that does not have this benefit list). As you can see, the combination in the first line has a 2.79% conversion and is winning this test. I have not gathered enough data to declare a statistically valid winner yet but we’re getting close. When the conversion rate comparison bar turns yellow (as it is here) this means that the results are still inconclusive. But we’re getting close!
Before you launch your first experiment you should have a clear vision on what you’re testing, why you’re testing, and what you hope to achieve through this testing. Setting up and conducting experiments in Website Optimizer is an efficient way to test various elements of your website. Now, make your plan on what you want to test, open your account, set up your experiments, and get to testing and increasing your conversion rate!