Let’s start with the basics – you have a website, and since you’re searching for more info on the web, you want to make it even more successful, or successful to begin with. 

And as we all know, a sign of a prosperous website is not in the number of daily visitors it attracts, it’s in the number of daily visitors that perform an action – proceed to the checkout with their shopping cart, sign up for the newsletter, buy a membership etc. 

How to Run an AB Test in Google Optimize

The number of users on your website who performed an action on a given month, divided by the number of users who visited your website during the same period and multiplied by 100 percent is called a conversion rate – and this is the number we want to increase to make your website more profitable. 

Say, you have 1000 website visitors a certain month, and 25 of them completed a macro conversion – purchased a monthly subscription, for example.

This means your website Conversion rate is:

25/1000 *100% =2,5%,

Which happens to be exactly the industry standard for any landing page. But if you want to increase your conversion rate, which you should, as some of the most converting websites score anywhere from 5,3% to 11,4%, you are going to need some quality AB testing.

So what is AB Testing? 

We, marketers, follow the same main rule the medics and doctors do, which goes – “Do no harm”, meaning we vow to improve your sales numbers, and never – deplete them. This is why when we propose a statement that says “your website might convert better if we change the carousel to a static image”, we first verify this hypothesis before implementing any changes. 

We start by creating two versions of the website or its element: 

  • version A –  a landing page with a carousel gallery
  • version B –  a landing page with a static image 

and after splitting the traffic equally between each landing page version, choose the one that got a higher conversion rate. This is website AB testing. To perform any tests, you’ll need a Conversion Rate Optimization specialist or the CRO-specialized agency, like our own, and the AB testing tools described below. 

But if you have some background on the subject or would simply like to perform the tests yourself, use our step-by-step AB test process guide from the next section. 

The Top Optimization and AB testing tools on the market are VWO, Convert, Optimizely, Crazy Egg, Optinmonster, Unbounce, Hotjar, PageSense and Google Optimize.

Google Optimize Description

Conversion Rate Optimization is all about the right use of resources and optimizing them to engage your users. And who’s best to help you with that if not the giant of web and search Google Analytics itself? 

Google Optimize is a native Google Optimization Platform, which has direct integration with Google Analytics, and allows you to run AB, multivariate tests, and redirect tests – all based on the previous data collected. You can create, adjust and change landing pages for your Google Ads campaigns right from your Google Optimize platform.

Overwhelmed with info?
Ditch all the instructions & contact
InsightWhale to run tests in Google Optimize for you 🙂

On top of that, Google Optimize also allows you to customize targeting and personalize your website experience according to your target audience needs.     

Advantages of Google Optimize

  • Offers a completely free plan 
  • Raw HTML, JavaScript, and Visual code editor
  • Easy implementation by adding a piece of code
  • AB, AB/n, multivariate, split URL, Server-side testing
  • Experiment preview, scheduling and management
  • Google BigQuery, Firebase, Google Ads, Accelerated Mobile Pages (AMP), and Google Analytics direct integrations which allow for fast data delivery and exchange

Disadvantages of Google Optimize

  • Unclear payment options: as a small to medium business, you can start free, but it’s unclear when your plan becomes paid 
  • As a large business or an enterprise, your payment plan will be determined by the sales team you’ll dial, costing up to $1,000/month
  • You can only run 3 experiments at the same time, using the free plan
  • You can only have 3 goals per one test, using the free plan
  • The number of Google Analytics multivariate tests is limited to 16 elements per test only
  • No real-time data receiving and reporting for the experiments

How to Run an AB Test in Google Optimize 

Step 1. Create an account

To start an AB test in Google Optimize, click this link, and begin achieving your optimization goals by creating a Google Optimize account.

Step 2. Make the required settings

Proceed with your Google Optimize account setup by creating your first experiment following the visual guide below. 

You’ll notice that you can create a variety of experiments, including redirect tests, multivariate tests, regular AB tests and personalization just like in the example image provided here:

Step 3. Link accounts

Link your newly-created Google Optimize account to the Google Analytics. 

Step 4. Make sure that analytics is set up

To make Google Optimize run on your website, you’ll need to set it up on this website using either one of the two methods, described under the links below:

Once you’re done or throughout the whole process don’t forget to make sure your Google Optimize is set up right using the instructions here.

You can implement your changes using two methods, described under the links below:

Step 5. Create test and implement variations

Create your first test and some first variations using either the visual editor or by adding a piece of code using the CSS code editor. 

This window, called the editor palette, opens on the lower right of your page and allows you to manipulate your variant however you want. Click on the “Edit element” button to edit your element’s HTML code, text, or even add a piece of JavaScript that runs when it loads. 

The lower part of the right window has all the control you need regarding the selected element – it’s location, dimensions, background, border, typography and layout.

In case you wish to have more control over your variants, use the other option – the CSS code editor. You can switch to the CSS code editor by pressing the symbol <> at the top of the app bar.

This version of the Google Optimize editor will allow you to implement all the CSS style changes while keeping all the non-CSS attributes like text and HTML intact.

Step 6. QA implemented changes 

The best way to perform Quality Assurance on the variation you’ve just created is to glance at the preview mode, which – a marketing hint – can also be used to preview an existing or a completed experiment. Preview mode allows you to see multiple pages you’d like to test the way your users would see it. 

To turn on the preview mode, follow these steps:

  1. On the experiment detail page, click Preview options.
  2. Then click Web Preview – and you’re done – or click Share Preview to create a link that can be shared within the team.

When using the Optimize extension for your Chrome browser, you can find and download here, you’ll find additional preview modes for Tablet and Mobile. When previewing a web page element you’re testing as a variant, you can get all the info on the container, experiment and the variant you are viewing by simply clicking on the Optimize extension button, where you can also turn off the preview if you want.

Step 7. Run the test

The best way to make sure your Google Optimize is running according to the plan is to run a small test personalization and employ installation diagnosis and the preview mode to monitor and verify your Google Analytics test is running appropriately. 

And here’s another little trick about using Google Optimize tests – you can schedule them! You can easily set the start and the end date for every experiment without you having to witness any of it. This can be extremely useful when it comes to certain period-driven campaigns during holidays or similar.   

All you have to do to start scheduling is visit this page for the step-by-step instructions right here.

Step 8. Analyze the results 

Once you’ve launched, ran and assessed your Google Optimize AB tests, it’s time to collect and analyze the results. 

There are two major ways you can analyze the results of your Google Optimize tests campaigns:

  1. Through Google Optimize reports, where you can see every little detail starting from the number of experiment sessions, dates they were carries out, improvements they brought etc. You can see the whole analytics report explained in the article here
  2. Through Google analytics reports, where you can access all the data, concerning your experiments, explained in the detail in the article here

There’s one more thing it’s important to mention when it comes to Google Optimize reports – they are delayed. According to the Google Analytics itself, the metrics are first processed by the Google Analytics and only then sent to the Google Optimize dashboard, which can take up to 12 hours from the end of the experiment. 

This is why you’ll generally see more info on your experiment in the Google Analytics dashboard described as an option 2 above, and “when you end an experiment, data accrued between the last push to Optimize and you ending the experiment will not be pushed to Optimize, but will be available in Analytics”.

Step 9. Enjoy your new Google Optimization 

This should be enough guidance for most websites, although if you experience troubles – don’t worry. According to Google itself, 

“If you notice page flicker while running Optimize, move the global site tag or Tag Manager installation higher in the <head> of the page. If flicker is still noticeable, install the anti-flicker snippet and customize it to suit your needs.”  

Other than that – you’re all done.

After all the instructions from above have been implemented, all the troubles – analyzed, and when you’re sure that variation you’ve tested wins – Congratulations! – you are ready to implement the changes you’ve tested to your website.


Without hesitation, AB testing is the single best idea you can employ to drive your conversion through the roof. Split testing allows you to get real feedback from the website visitors, which allows you to make data-driven decisions and be sure they will result in more conversions from your visitors, and therefore – more sales.

When it comes to using Google Optimize as your top optimization tool, you should evaluate your conversion needs before implementing it first. If you are a small to large companies that just starts optimizing and doesn’t need more than 3 tests at a time and 16 elements to test, you can definitely start with the free plan Google Optimize has to offer. 

But if you are a small to medium company, whose needs well exceed the limits of the free plan proposed by Google Optimize, you may consider other choices. The paid plan can cost up to $1,000 a month, so if you’re not looking to invest into conversion optimization at this stage that heavily, it’s better to turn to other, more affordable AB testing tools.

In any case, make sure you consult a CRO-agency or a specialist who will help you decide which optimization option is best for you.