A/B Testing With Google Optimize

Share

AB Testing With Google Optimize

A/B testing is a great way to improve your website performance. You could imagine that it would be excruciating to make small changes and try to decipher the impact this has had in your analytics account.

This is where Google Optimize comes in.

Improving conversions by AB Testing With Google Optimize

What Google Optimize allows you to do, is to make a small change (without even touching your website coding) and sends half of your traffic to the original version and the new version.

They then monitor this campaign on your behalf and report back on how a given conversion is performing on each version, this conversion may be bounce rate, time on page, add to carts, contact form submissions or anything else you might want to measure.

How To Set Up Google Optimize For AB Testing

Step #1 Go to the Google Optimize website and sign up.

Google Optimize Front Page

This is the first page you will land on after signing up, click on create experience to get started.

Google Optimize Step 2

Give the campaign a name and enter the URL that you would like to test.

You want to select A/B testing for this experiment, what this will allow you to do is make adjustments on a page within the Google Optimize system rather than have to create two separate pages.

Once you’ve completed this, you will be directed into the campaign settings.

Step #2 Before putting together your campaign, you will need to link your Google Analytics account, they use this to measure your conversions.

On the right of your experiment there will be a box that says you haven’t connected your Google Analytics account yet.

Connecting Google Optimize To Google Analytics

Click on Go to container page to get this started, this will redirect you back to the main page.

On the right will be a bar saying that you have not finished setting up your ‘container’.

Scroll down on this sidebar and there will be an option to LINK PROPERTY.

Click on this, if you are logged into the correct Google Analytics account then this process should be fairly automatic.

Step #3 Now you will be asked whether or not you want to add the Optimize snippet to your website right now. Might as well get it done and then you’re all set to get your experiment going.

Google will give you 3 steps, which is basically just to add one line of code into your existing Google Analytics snippet.

ga('require', 'GTM-XXXXXXX');

On the next screen they also ask you to add the anti flicker code to your website, just before the analytics script.

This anti flicker script will prevent the user from seeing your changes move around on your website.

Google Optimize Code

The above is the final code I added to my wordpress website.

Doesn’t really matter how you add it, as long as it’s as high in your head section you can manage.

Now that’s out of the way, your all set up and ready to get an experiment started.

Click done, then click on your campaign you made before to get back into the settings.

Step #4 Click on Create variant to the bottom right of the bar that says Original.

This will be the alternative that you are wanting to test, whether its different text or a different colour so name it according to what you are wanting to test.

Now if you click on the variant you just created, it will ask you to add the chrome extension. Go ahead and install it.

Now when you return to the variant, if you click on it again it will open the Optimize editor.

Google Optimize Editor

This editor allows you to make a huge amount of changes, it’s a neat tool.

You can try out a number of variants such as

  • The text you use
  • Sales copy
  • Any colours
  • Moving elements around
  • The sizes of anything

Pretty much you can make any changes that you would otherwise manage with CSS.

It’s best to only make one change per experiment so you are 100% sure what is impacting on your results.

Once you are done making changes click on DONE in the top right.

This will return you back to the campaign page.

Google Optimize Variants

You will notice that on your variants they show a %. This is the amount of traffic that is divided between your variants.

My variants are showing 50%, so this means that 50% of the traffic that lands on my page will go to the original page and 50% will go to the variant.

You can adjust these if you want by clicking on the number.

Step #5 Scroll down a little further to the objectives section, if it is asking you to do so, connect this campaign to a Google Analytics View.

Now you should have an option available to ADD EXPERIMENT OBJECTIVE.

You will have a choice between choosing from a list and custom. The list will be a few standard metrics to measure that are provided by Google Optimize.

The custom options will allow you to track events which will be fed from Google Analytics. This allows you to track anything you want if you are familiar with event tracking.

For simplicity, we will select Session Duration. This will measure session duration between the original version and the variant version.

Step #6 If you move over to the TARGETING  tab this lets us decide when these Google Optimize changes are triggered.

Google Optimize Triggers

If you only want to test a portion of your traffic, you can adjust the slider under “percentage of visitors to target”. This is pretty self explanatory, it will limit the amount these optimized pages are shown.

The main thing you want to look at is the URL matches section.

This will determine which pages the optimization is active on.

Now, it’s worth clearing up that the URL you entered earlier is simply the URL you want to test the editor on, so when you are making these variant adjustments it will use the earlier URL as the editor page.

This section is for defining when the optimizations are active.

So, if you wanted to have it displayed on a number of different product pages you could set it here.

Google Optimize URL

Now, you’re done!

All you have to do is scroll back up to the top and the grey bar will say “your experiment is ready to start”.

If you would like, you can set up a schedule so that it starts and stops according to whatever dates you would like.

When this is done, click on start experiment and you’re all done!

Step #7 Now you just need to monitor the campaign to see which is the victor, up the very top of your campaign page will be an option for reporting.

Let your campaigns run for at least a few weeks, longer if you don’t have much traffic.

Google will let you know if something is a clear winner, then you just go in and make the changes permanent!

If you have any issues, i’d be happy to help! Shoot me an email at chris@christopherwd.com


Share

Leave a Reply

avatar
  Subscribe  
Notify of