What Is A/B Testing?
A/B testing can also be known as split testing or bucket testing (lesser used). These three terms are thrown around interchangeably but all hold the same meaning. In short, AB testing is a way to compare two versions of the same webpage (with a small change on one of them) against each other. The tests performed are a randomized experiment with two variants; A & B. After the A/B test is complete, you’re able to see which version performed better and then make the necessary changes to the live website. There are many softwares and platforms that are made solely for A/B testing that have easy reporting features, enabling you to see what has worked or what hasn’t (which is equally as important) and thus increasing your conversion rate.
A/B testing allows companies and their marketing agencies to make data-driven decisions and to make informed conversion optimization decisions. It’s important to realize that each vertical in business is different and what has worked for one company will not be a blanket approach. This is why A/B testing is so important for each business.
How Does A/B Testing Work?
When creating an A/B test, you are essentially taking a web page and copying it to create a second version of the same page. However, this second page will have a small change within it. This change will be the variant that you’re testing out against the original page. This change can be as simple as a single headline or colour of a button, but it can also be a complete redesign of the page (redirection/multivariate testing).
Once the test is live, half of your traffic will be shown the original version of the page (known as the control) and the other half of the traffic will be shown the modified version of the page (the variant). Whichever platform you are using for your A/B tests will then start measuring and analyzing engagement rates within both of the webpages. You can then determine whether changing the visuals/functionality of the webpage had a positive, negative, or no effect on user behaviour.
Each indication is equally as important, as you will know what resonates and what does not resonate with your users. Our preferred A/B testing platform at Consultus Digital is Google Optimize. Google Optimize allows for such easy element modification, that you don’t even need a developer!
Should You Be A/B Testing?
Yes. It doesn’t matter what vertical you are in or how small your business is, you should always be A/B testing.
A/B testing allows businesses and marketing agencies to make data-driven decisions to which changes should be made to webpages in order to alter user experiences, while also collecting useful data. It allows businesses to understand why certain elements of experiences impact user behaviour more so than others. As mentioned, it’s equally important to know what doesn’t work.
After completing a single test, test again! Don’t just run one test and stop. You can always improve your website, especially because user behaviour will continue to change. Continue testing and improving the conversion rate. However, it’s important to note that testing one change on a single page at a time helps pinpoint which changes had a positive or negative effect on user behaviour, and which ones did not. There’s a never-ending list of changes that can be made to web pages, albeit some more effective than others.
Steps For A/B Testing
Determine The Conversion To Improve
When trying to determine a change for a website, it’s important to dig into any data that you have collected in the past. Run through Google Analytics to look for pages with a low conversion rate, or review any heat maps or screen recordings to see where drop-offs on the website occur. The pages with the lower conversion rates are where you want to focus your efforts first.
Hypothesize The Change
Now that you have a page that you want to focus your A/B testing on, you need to determine the exact variable that needs to be tested on said page that will align with your overall goal. What on the page would lead to higher conversions from that page? For example; if on that page your call to action button is dark and halfway down the page, moving it up and changing the button to a bright colour would be a good hypothesis that would lead to increased conversions.
Since you now know what the variable for testing will be, each of the variations of the website needs to be created. This can easily be done through the platform that you will be using for A/B tests. There’s normally a step by step guide to follow. If you have any trouble, our SEO experts can help!
Run The Experiment
After the variations are created within the platform, ensure that your goals are linked to the experiment in order to track results properly. Goals can normally be linked via your Google Analytics conversions. Once this is complete, push the experiment live!
Measure The Results
The platform will normally notify you after a threshold of traffic or number of conversions have been reached and therefore the experiment will be complete. Each platform should have a reporting feature that reviews the amount of traffic to each version of the page, along with the conversions and conversion rates for each.
5 Testing Examples
CTA Button Language
Try simply changing the copy on your call to action button. Certain copy entices users more so than others. ‘Download Now’ may work better than ‘Learn More’ as it is asking the user to perform a timely action.
CTA Button Colour
Dark button colours blend into websites, especially when it’s a call to action. You want to ensure that the buttons are bright and stand out from the rest of the website to get users to click on them and perform an action.
Banner Image / Video
If you’re marketing to 45-year-old males but your website header image has an image of a young female, you’d want to try testing out a different image to see if that helps lead to additional conversions. Users like to see images of people who mirror themselves on your website.
If your call to action button is at the bottom of the page, users generally don’t scroll down very far. Try moving the placement of the button to the top of the page in the header image.
Looking at the copy on the web page, try testing out 500 words vs. 1000 words. Depending on the vertical, less information on the page entices users to fill out the form in order to get more information. For some, more information for bigger purchase items helps promote trust and therefore the conversion.
A/B testing is important for any business, as it generates useful data that can improve conversion rates and drive more leads. It also gives you more insight into your consumers and what they find to be valuable. In terms of marketing, it also aids with SEO, as Google encourages different forms of testing.