Whether you’re looking to increase web sales, get a higher number of subscribers to your site, or simply measure visit duration and click throughs, A/B testing with Google Analytics Content Experiments can help you reach your business goals. This simple tool is one of the best ways to make use of your Google Analytics account, and we’ll show you how to use it today.
Although it sounds highly technical, A/B testing is actually very simple. Think of it as a three-step process:
This easy, three-step process is a vital part of reaching your goals.
Consider a webpage that has a 25% rate of conversion, meaning that for every 100 visitors, 25 signup to receive your weekly email update. But what if you create an experimental version of the page and using Google Analytics Experiments find out that the page variant yields a 50% conversion rate? A huge improvement.
Now here’s where it gets interesting.
That same kind of testing can be applied to your sales page. Let’s pretend that your sales page has a 5% conversion rate: i.e., for every 100 visits to your sales page, you make 5 sales. But what if your secondary, experimental page shows a conversion rate of 15%? You’d probably immediately implement the changes you made to your secondary page directly to your original page. Fortunately, the Experiments algorithm will send more of your visitors to the winning page automatically, so you can continue to let your experiment run. Experiments typically run for 2 weeks, however if you get statistical accuracy, you can finish early.
But the smart business owner doesn’t stop there. Because Google Analytics allows testing for up to five variations of a page, good webmasters will continue experimenting.
Before we get ahead of ourselves, let’s talk about how you can get started with Google Analytics Experiments on your site. It’s easier than you think.
After you’ve got two different versions of your web page ready to be tested, follow these simple steps.
Click “Create Experiment” and enter the URL of the original webpage (the webpage you are testing your variants against) and click “Start Experimenting”.
Be very descriptive when choosing the Experiment Name, it is very likely you will create many different experiments and if the names are vague it will be confusing. Also be careful, if you have a running experiment that is similar to a new experiment and you attempt to run them at the same time, it causes really stranger errors and both experiments will become unusable.
Notice we selected an Goal/Conversion that was created previously for the Objective. Its possible to use other metrics for the objective. There is also the option to create a new Goal here, but we have found it works better to create the Goals before creating the Experiment. If you have not set up Goals already please visit our article here and set those up before continuing.
The other settings are optional. Depending on the requirements these change how the experiment will run.
Here, the Original Page is auto-populated from Step 1, now its time to add the Variations.
Again, make the names for the pages as descriptive as possible, so the reports are easy to read. Notice at this point you can Save the experiment for later.
Here we can review the configuration and choose to copy the code “Manually insert the code” or email the code to the webmaster. This code snippet goes on the main page (Original Page), it’s not required on the variant pages.
We click “Manually Insert the code”, meaning we are going to copy this code and add it just after the <head> tag in the HTML, as directed.
Here we can pass along a very helpful tip. Sometimes, to make our Experiment code we copy from Google work, we have to add this snippet above their code:
_udn = “www.abundancethebook.com”;
Before clicking Next Step, you have to add all the code to the Original Page. Clicking Next Step will check the Original for the code, and make certain the Experiment will work properly. If there are errors, this validation step will attempt to help fix it.
When the Experiment code is verified and validated as working, you get the option to Start the Experiment now, or save this Experiment for later. This is helpful because you can plan a bunch of experiments and create them before actually starting them. Please keep in mind if 2 experiments with the same Original page are running at the same time, it will cause errors.
The easiest way to get the code on to your page is by using the Google Content Experiment Plugin for WordPress. Simply install the plugin, click on page editor, and insert the code. When you click “Next,” your code will be validated. You’ll then be prompted to review your content and review and start your experiment.
You can check your Experiment Reports in Google Analytics to see the status of your experiment, your page performance, and notable changes in conversion rates.
Pay close attention to which page brings in more subscribers and greater sales: significant increases to the experiment page are changes you may want to implement into your original page later.
Other than the fact that it’s easy and completely user-friendly, Google Analytics lets you take control of your webpage so that you can reach your business goals.
Even small variations to your page can increase subscribers and sales, so remember to keep testing! Leave us a comment below if you have some tips to share and don’t forget to subscribe to get our latest articles delivered to your inbox!