Experiments allow you to test how different versions of your pages work to get visitors to meet specific objectives. You can test up to five variations of a page to see which one generates the most conversions.
The Content Experiments feature is a different method from standard A/B and multivariate testing, as it is more A/B/N.
It is not about testing only two versions of a page, as in A/B testing, nor about testing many combinations of components on a single page, as in multivariate testing. Rather, it involves testing up to five full versions of a single page that are displayed to visitors via different URLs.
With Content Experiments, you can:
- Compare how different web pages perform with a random sample of visitors.
- Specify what percentage of visitors is included in the experiment.
- Choose the type of objective you want to test.
- Get email updates on the status of your experiment.
To understand it better, we will give an example of using experiments to improve a business. Let's say you have a website that sells domestic cleaning services. It offers basic cleaning, deep cleaning and general cleaning. General cleaning is the most profitable of the three, so you want to get more users to sign up for that option.
Most visitors enter the home page, so that's the first page you'll want to try. For the experiment, you need to make several new versions of the web page: one with the general cleanup highlighted in a big red title; another with expanded information on the advantages of general cleaning; and another with an icon next to the link to hire it.
After setting up and starting the experiment, a random sample of visitors views the different pages, including the original home page, and we just have to wait to see which page gets the highest percentage of visitors to hire the general cleaning. Once you know which page converts the most, you can make it the default page for all visitors.
