Guides & tutorials
Guides & tutorials
Google Optimize is a free A/B testing tool that allows you to test which variation of your website works best, but also check which kind of form converts better in your case.
This is a good method to discover if your forms are adding too much friction to your user experience, and to try different options to improve them. Do you want to learn how? Just follow these 8 steps to easily start running A/B tests.
If you haven't already done it, simply create a new account at optimize.google.com. You can get the snippet from 'Container settings' section. Just copy this line of code and paste it in your website. You can also check official video to see how to do it.
You will also need to connect your Google Analytics account, in order to collect your A/B testing data, and install the Chrome extension to use the visual inline editor. Simply click on the respective buttons to set up these features.
Go to main page, click on the container where you want to create the test and, then, click on 'Create experience' button, on the upper right hand corner.
Give your experiment a name, and select the 'A/B test' option at the bottom of the menu, as you can see in the picture below.
Once it is created, we simply have to set up this test, based on what we want to measure. In this case, we would like to check which 2nd auth factor converts better.
Open your new experiment to open set up options, and click on 'Add variant' button to create a new one. Experiment will be saved as a draft during its configuration.
To run an A/B test, evidently we need to create two variants of the same page, which we will call, for example, 'Original' (the page which is currently published, that sends one-time passwords by SMS) and 'Email to OTP' (the one we are going to compare with).
As you will see, the second variant includes a change log (now 0), so next step is just to make the changes whose results we would like to compare with the original one.
Click on 'Edit' button to open the visual editor, remember that it's needed to install the plugin before, and make the proper changes. To change the form that will be executed on this variant, click on your 'call to action' button > 'Edit element' > 'Edit HTML'.
Using the HTML edition view, you just need to change the form ID, as you can see in the picture below. If your website isn't already using Arengu forms, you can check how to do it in our documentation, simply using method 2.
To better distinguish at a glance which of the variants are you seeing, we recommend you to change a little detail on it. For example, the color of the button. Once you have included form ID, click on 'Done' button to save variant changes.
To get the form ID to show on each variant, just go to your form's edition view on Arengu, and then 'Settings' section. Here you can find your form's ID, to use it on the proper variant of your A/B test experiment.
You can also replace form ID on 'Destination' field, on Google Optimize, but it would be easier to edit on the HTML edition view we explained on the previous section.
The next step is simply to indicate in which view of our Google Analytics account we want to save the data from this A/B test. Go to 'Measurement and objectives' section, click on edit icon, and just select the view from drop down, as you can see in the picture.
In the same section, you can include a custom objective or get one from a predefined list. Goals created in your associated Google Analytics view, will show up here. For this experiment, we are going to select signups to measure success.
To add custom goals to the experiment, just connect your Google Tag Manager account.
Once everything is set up, you will see that the 'Start' button is activated in the upper bar. Juts set the start and end dates, and click on 'Start' button to launch it.
From now on, you can check the evolution of your experiment on this page, and also edit and stop it when you have enough data.
Thanks to how easy it is to make and publish changes with arengu, you can test and improve your onboarding and user verification and authentication processes in a very agile way, to quickly find the option that best suits your real users.