What is Google Optimize?
Google optimize helps in testing and delivering optimal website experience. Optimize is built on Google Analytics, which means we can use the customer insights which mean we’ve already configured within Analytics to test against business metrics that matter — such as ad revenue and e-commerce transactions — without any additional development/coding work.
Some feature are as follows
- Put customer insights to work: Native integration with Google Analytics provides us with the rich behavioral insights that we need to test and deliver optimized customer experiences.
- Fit the moment to the customer: Advanced experiment targeting means we can deploy the right experience to customers at the right time.
- Easy website testing: Optimize makes it easy for you to create A/B, multivariate, or redirect tests to change just about anything on your site without any additional coding work and then quickly determine the impact on customer behavior.
- Quick Set-up: Quickly implement Optimize across your entire website and start running website tests in no time.
How to configure Google Optimize?
- First, we need to navigate to https://www.google.com/analytics/optimize/
2.Click on Sign up free as shown in the above picture
3.Install the below script inof the page as shown in the below image
4.Once script is installed. We need to install Google optimize add-on in the chrome browser
5.The following webpage will appear once login into the optimize page
How to create an experiment?
- Click on Create Experiment button
2.Give a name to your experiment
3.Choose your website
4.Choose experiment type
- A/B test
- Multivariate test
- Redirect test
5.Then click on “create” at the Right hand top corner
6.Select Start date and End date of Experiment
7.Choose Variant of web page
9.Choose target customer
10.Click on save
To check how any experiment is doing, we need to check its report
Click on Reporting
Check for performance based on experiment session.Check for objectives based on pageview and bounce rate
If the hypothesis is proved with the available data then, use the variant which performed well in the experiment