How to use Google Analytics to Split Test Different Pages
We all know how important it is to split test, optimize and all that good stuff. It is vital for any marketing strategy. There’s absolutely no reason to waste all that effort driving visitors to a website not knowing what piece of content they respond best to.
Well, I can’t think of a better tool to use when it comes to split testing different pages on your website.
Not only is Google Analytics free, it is the easiest option when it comes to testing different page variations – such as designs, headlines, content and so on. 🙂
Before you even create a split testing campaign, you’ll need to set up Google analytics for your website.
What You Will Discover
- Importance of split testing pages
- How to set it up within Google Analytics
- Testing different variations
- Where to look for the statistics
Thanks For Sharing!
Here’s the PDF: Download the PDF
[su_heading size=”24″ margin=”40″]Setting up A/B Split test with Google Analytics[/su_heading]
First, you will need to create the page you want to test the original against (That’s the whole point of split testing). The page to be tested against the original is sometimes referred to as the variation page.[su_frame][/su_frame] Although the heading says A/B testing, you are not bound to only testing 2 different pages. You can test as many different pages as you like.
However, without a high enough volume of traffic, the test might be spread too thinly across the different pages. I also recommend you start by split testing one variable using A/B split testing only (as opposed to A/B/C/D/E split testing for example) because it can get confusing quite quickly. Also make sure you have all pages published before you test anything!
Once you have created all of the pages to be tested and have made them publicly accessible, head over to ‘Experiments‘ just under the ‘Behavior’ section in Google Analytics.
Google likes to call their tests ‘Content Experiments‘ and that makes sense, but really, it’s just a fancy way of saying ‘split testing’. 🙂
Since this will be the first time you are creating an experiment, you will see a blank table with no data.
So create a ‘New Experiment‘.
Next, you will be prompt to fill out some of the details for your test.
To start off with, you have to name of the experiment. Think of something really fancy like ‘Decrease Bounce Rate’… 😛
Under that is the objective for this experiment. Think of objectives like creating goals within Google Analytics. In this case, you have the choice of picking a templated metric or create a new one. In my example, I chose “Bounces” under ‘Select a metric‘.[su_frame][/su_frame]
The really cool thing about creating these tests is that you can choose to test a certain percentage of the traffic you want to test and send that to a specific page.
For example, if you only wanted to test half of the traffic on the split testing campaign, 50% of the traffic will land on the page being experimented on, whereas the other 50% will always land on the original page (which is your controlled variable).
You will see more options by clicking ‘Advanced Option‘, which to me, aren’t really ‘advanced’, so don’t be afraid to check them out.
The 3 advanced options include whether you want to distribute all of the traffic evenly, how long you want to run the test for and it also allows you to specify a confidence threshold.
The first 2 options are self-explanatory.
The last option indicates how confident you want Google Analytics to be when it comes to determining a winning page.
The higher the percentage, the more definite and accurate the results. So at a confidence interval of 99.5%, it means that it is 99.5% confident that the one page is out performing the other.
All that’s left now is to click on the “Next Step” button.
Here is where you want all of your links ready to lock and load!
You will need to add in the URL of your original page. You can name the page whatever you please, but generally I like to keep it as ‘original’. (Simple!)
Under that, you will add the first of the variation pages. If you want to add more pages, just simply click “Add variation” and another entry will pop up.
I suggest using easy tags so that it’s not such a headache deciphering your data later on. In my example, my end tag for the original link is “/top-3-amazing-dog-tricks” and for the variation link I used “/top-3-amazing-dog-tricks1“. Simple right? 🙂[su_frame][/su_frame]
After clicking ‘Next Step‘, you will be given a code that consists of a bunch of random numbers and letters. This code will need to be added to your WordPress site header. This is probably the most difficult part of it all, but don’t worry! I got you covered. 🙂
Let’s start by clicking on ‘Manually insert the code‘ and select + copy that fine piece of script code.[su_frame][/su_frame]
You will have to paste this code into your original page right under the <header> tag. If you don’t know how to do that, it’s OK! Good thing there’s a plugin that can simply do that for you!
This plugin is called “Google Content Experiments” and once installed and activated, there will be an option to add experimental codes. This is found under the ‘Content‘ tab when you “Edit” one of your posts.
Once you have added your code into your original page/post, all that’s left is to check whether or not Google Analytics picks it up.
After you get the green light, you are ready to start the experiment!
Visitors will start being redirected to all of the experimental pages. Some will see the original, some will see the variation page. Google Analytics will let you know how your audience responds to your page.
You can now test all sorts of pages and any crazy things you can come up with.
Have fun! 🙂
Note: It will take Google Analytics a few days to collect data before it shows you any kind of result. Just know that it’s working around the clock behind the scenes to do all this intense data collection for us. 🙂
If you have any questions, please feel free to drop a comment below. If this tute helped you out, share the love by clicking on one of the colourful buttons below.