Google Optimize is a tool used to run “experiments” on websites and customize content in order to target different audiences. The goal is to optimize websites for user experience by running controlled tests of different variants of the same content to choose the one that works best, i.e. to create A/B tests.
Google Optimize is part of Google’s suite of tools aimed at improving companies’ digital marketing efforts.
Google Optimize is part of Google’s suite of digital marketing tools “Google 360”along with Google Analytics, Google Attribution, Google Audience Center, Google Tag Manager and Google Surveys.
To start using Google Optimize, it will be necessary to install the tracking codes for both Google Analytics and Optimize (as they sync and work together) on the website you want to optimize. You will also need to use the Google Chrome browser and the Optimize extension.
What kind of experiments can Google Optimize do?
-A/B test: This is the most basic type of experiment, as it involves creating two versions of the same page, one of them with a different element and analyzing which of the two works better. It is an experiment that can be used to find out how a specific design change would affect, for example, changing the size of the typography or the header image.
-Redirect test: It works similarly to the previous one, but instead of making the change from Google Optimize, it allows the user to use a different URL. This is an experiment designed to test more than one change, especially designed for a format change, for example, to test how a website redesign would work.
-Multivariate Test: Allows you to analyze how various changes would work. For example, to test different fonts or image sizes. For example, if you want to change the size of the headers, but also the color of the links, instead of doing several A/B Tests, you can perform a Multivariable with all possible combinations to analyze which one will work best.
-Multipage Test: Works like an A/B test, but create a variant on more than one page of the site to improve the test results. For example, an ecommerce wants to know if “add to cart” or “add to cart” works better, with this test you can change several product pages to see which version works better.
How to create an experiment in Google Optimize?
Then, you need to follow some very simple steps to start with the first tests:
-Define the test parameters: You should identify all those aspects of the website that you want to improve and make hypotheses, for example, “increasing the font size will reduce the bounce rate”. It is important to select a metric to have something to measure.
-Start: To do this, from Google Optimize you must click on “start experiment”, which will open the experiment configurator. You must give a name to the experiment, enter a URL and the type of experiment that best fits among the options offered: A/B test, redirection test or multivariate test.
-Original version: The original page will then be automatically created which will serve as the control version for performance comparison.
–Select target: Google offers a series of predefined goals (page views, session duration, bounce rate, etc.), but it also allows you to configure custom goals. In this section it is also possible to configure the number of visitors to which the experiment will be shown, as well as the distribution of traffic among the variants.
-Preview and launch test: Finally, just run a preview test, to make sure everything is working properly, and launch the experiment. Once launched, the results should be accessed periodically to analyze the performance of the experiment.