Google Optimize enables you to run and measure experiments/ use cases/ test types on your website based on a hypothesis. These experiments are often referred to as experiences or personalization based on the experiment context. Basically, Google Optimize allows you to:
- Test a hypothesis by running website experiments.
- Deploy the leading variant permanently.
- Personalize your website to create the best experience for your visitors.
Google Optimize Experiment Elements
A Google Optimize Experiment has the following basic elements
There will be two pages in the experiment as you can see in the above screenshot. One of the pages will be the original page and second page also known as the variant page will be the page with changes.
Weightage decides in percentage in which you want to split the traffic in between the original and the variant pages. Ideally, between two pages, it becomes split testing and the equal weighting of 50-50 percent is given to both of the pages. If there are more pages involved then weighting is shifted accordingly.
Editor page & Page Targeting
Editor page is where you make changes on a Live page to test your hypothesis. While Page targeting is the set of pages where you want to run the experiment on.
The objective is the goal that you want to test. These goals can also be a Google Analytics Goal which you might want to try. If you want to test some event specifically, I would suggest setting up an event first through GA or GTM, then set up a goal based on an event which can finally be tested
Use Case 1: Related Post BG Color
The idea of this experiment was to test the CTR (Click Through Rate) of the related posts or suggested posts which are shown at the end or in the middle of a single post based on their background color. Here is a screenshot of the related posts which would earlier appear on a single post.
Because the background is white here and also, the single post background was white, my hypothesis was that it looked more like a series of pictures with some captions that suggested posts. So, I decided to change the background color of the suggested posts to light grey.
You can clearly see that variant 1 i.e. variant with a light grey background has far higher CTR compared to the baseline i.e. page with the white background on the related posts.
Use Case 2: Table of Page Content on each page
The idea of this experiment was to test the bounce rate of a page when each page will have a table of page content. Earlier, my individual blogs will not have a table of page content. But I was inspired by the idea of a Wikipedia page and I thought that it would be nice to have a table of page content from the user experience point of view. Here is how my pages look like now with the new table of content embedded in each article:
Below is the screenshot of the result. You can clearly see that variant 1 i.e. page with a table of page content has a very high bounce rate compared to a page without any table of content. However, I still do not care for bounce rate as for me user experience matters more.
Use Case 3: Larger Font Size Across Devices
The idea of this experiment was to test the font size of a page. Recently I have come across a number of blogs where the font size will usually be larger. From a user experience point of view, it is a good idea to increase the font size and switch over to a bigger font size for better engagement.
However, I also realized that the idea of a bigger font size could have been tested better if different experiments were run for different devices. After realizing that I started a separate experiment to test for session and bounce rate on the mobile device when the font size is a little bigger than the regular font size.
Below is the screenshot of the result. You can clearly see that variant 1 i.e. page with larger font size has a lower session duration compared to the original page.
Use Case 4: Main Blog Page White Background
The idea of this experiment was to test if there is an increase in the session duration and bounce rate with a different background on the main page of the website. Currently, on my travel blog, the background is greyish. The idea is to test the engagement factors when the background is white.
There was not any clear winner in this case. But as you can see below. The main page with white background has a better bounce rate but also a lower session duration chances. So, I decided not to continue with the white background idea rather I am going to continue with the grey current background.
Use Case 5: Larger font on Mobile Devices
The idea of this experiment was to test if there is an increase in the session duration and a decrease in the bounce rate on a mobile device if I make the font size 10-20% larger on my current travel blog. Well, Google Optimize could not find a clearer winner as you can see in the screenshot below.
There was not any clear winner in this case. But as you can see below that larger font size on the mobile device has a lower probability for beating the base line.
- Changing the background of the related posts increased the engagement on my website
- Introducing a table of content decreased the session duration and also increased the bounce rate but I am still going to keep it as it helps the user in finding out the right content through the table of content
- Larger font size across the device or only on the mobile device had not much of a positive impact in terms of engagement on my website
- Changing the main background of my blog also did not do any magic in terms of engagement
If you have some similar idea, do let me know in the comments below.