How to A/B Test With Google Optimize
You’ve decided to revamp your company website. It’s been long overdue, and your graphic designer has come up with some great concepts.
But which one do you choose? How do you know which one will work best?
It sounds like you need to run some A/B testing! Lucky for you, there is a brand new tool out there that makes this approach smooth and efficient.
Google Optimize is fresh out of its Beta phase and is ready to transform your business immediately. The tool is free to use, but how exactly do you set it up for A/B testing?
Let’s take a look together.
Let Google Optimization Enhance Your Performance
To run A/B tests, you need some changes and tweaks ready to roll out. Think small.
You want more customers to sign up to receive your newsletter and email marketing shots. So play around with the color and positioning of your subscribe button. Set three or four different positions and colors.
Now you are ready to test.
1. Create an Experiment
Google Optimize calls your A/B tests “experiments,” because, well, that’s kind of what they are!
As soon as you log in, you will see the link to your experiments page. Just click create, and you will be on your way.
An experiment is set up in three easy steps.
- Enter a name / description
- Enter your URL
- Select the Experiment Type. In this case, A/B Test
2. Set Your Testing Parameters
When using Google Optimize, you can set several levels of parameters, which will allow you to leverage the tool to the best of its capabilities.
Your parameters will include deciding:
- Overall Objective
- The type of visitor you wish to target
- Weighting – towards one particular variant or another.
- When to target people
- Targeting Rules
With your parameters set against your base experiment, it is time to move to the next stage.
3. Variants Deliver the Best Results
With testing, you want to understand what is going to work for you and your customers. Variants are exactly what they say they are. Variations to the primary experiment!
To go back to the ‘subscribe’ button. You have your experiment set to your main site. You then create three variants showing the three different positions of the button.
Each variant will be ‘live’ for customers at the same time, and unless you have played with the weighting, there is an even chance of anybody being sent to either one.
Now it is Up to Your Users to Deliver you Results
All you need to do now is hit Start, and your test will begin.
There is no hard and fast rule on how long a test should run. It all depends on the number of changes you are making, and the degree of change from one variant to another. Google recommends at least two weeks, which is a good timeframe.
All you need to do is use the reporting and analytics tools Google optimize provides you, and watch the data build up.
By the time your tests are finished, you will know the answer to your questions and will be able to move the subscribe button to the position that proved the most successful.
With continuous releases of new digital products, A/B testing should be common place.
Reports show that 61% of companies run approximately one test a week.
Stay ahead of your competition by ramping up your A/B testing.
Start using google optimize today to push your website to the next level!
WordPress SEO plugin
SEO Booster PRO
Learn more about your website SEO and do something about it.
Developed, tested, tweaked and improved since 2008.
I have changed my mind. The Shortpixel service is my new favorite image optimization tool for WordPress websites. I have been using EWWW Image Optimizer for a long time, and I still like it, but Shortpixel has become my new favorite image optimization...
SEO Booster Free version Not too long ago I silently released SEO Booster on wordpress.org - Yipee! I did so without a big announcement because I wanted to catch and fix bugs without too many users affected. Since the initial release, I have released...
Are you unsure of how to effectively reach potential clients and readers online? No fear. Here are the most vital digital marketing concepts you need to know.