In the online world, competition is fierce. In addition, the behavior of visitors is constantly changing. By using A/B testing, you can be sure that you won't miss any opportunities for conversion. But how do you get started with A/B testing? In this blog post you will discover.
What is an A/B test?
An A/B test or split test is a test where you pit two variants of a web page (or specific parts of a page) against each other to see which one is more effective. You present the unmodified version of your page, variant A, to a group we call the control group. The modified version is then called variant B. (You could also add variants C, D, and so on.)
The versions receive visitors one by one to achieve a representative comparison, and after a certain period you evaluate the results.
By always building on the most successful variant of two tested pages (or parts of them), you increase the conversion ratio of the page step by step. An A/B test is therefore an indispensable marketing tool for anyone striving for online success. As specialists in converting websites, we regularly use A/B tests.
How does an A/B test work?
An A/B test starts with a hypothesis: a suspicion you have, based on statistics from Google Analytics, or inspired by a blog post about online marketing. For example, you may suspect that an orange call-to-action button will lead to more clicks than your current, blue button.
Then you determine how long you will test this. It is advisable to take the same period for each test. If you choose a week, test for example from Monday to Sunday. If visits to your website are seasonal, avoid the weeks with big peaks or troughs in visitor numbers. At 2manydots, the starting point of a test is at least one month.
After configuring and activating an A/B test, visitors get to see a variant of your website in turn. The first visitor sees version A, the second visitor variant B, the third visitor variant A and so on.
It is important that in an A/B test you only test one small change, for example the color of a button or a certain header above a text. If you test more things at the same time, the difference in conversion cannot be directly linked to a specific change.
At the end of the experiment, you see the statistics for the control group compared to the statistics for the variant(s). You can then see to what extent your hypothesis was correct and which direction you should go in terms of UX design or content.
Available tools for A/B testing
There seem to be more and more tools for conducting A/B tests. They range from free to a hefty monthly fee and between them there are big differences in the bells and whistles. We won't go very deep into these tools in this article, but we will briefly mention a few for the sake of completeness:
- One of the best known tools is Optimizely. It is used by large organizations and has a corresponding price tag. With Optimizely you can run tests based on cookies, on demographics and location. It also allows you to test on iOS and Android.
- VWO is another popular tool for A/B testing. Short for Visual Website Optimizer, this is a very user-friendly tool with a friendlier price. For smaller companies, this is an attractive option.
- Crazy Egg is an A/B testing tool that also allows you to create heatmaps. The price tag grows with the number of visitors you have.
- Google Optimize is a free tool with, as we have come to expect from Google, many features. This tool is perfect for those who want to take their first steps in A/B testing.
In this blog post, we focus on the latter tool: Google Optimize. You'll learn what this tool can do and why it lends itself so well to WordPress sites.
Why choose Google Optimize?
Google Optimize is approachable on the one hand due to the fact that this tool is free. There is also a premium version for enterprise-sized organizations: Optimize 360. In the early stages, the free version will certainly suffice because it allows you to run up to five experiments simultaneously.
On the other hand, this free A/B software can be overwhelming because it has a huge number of functions and that in turn requires the necessary technical knowledge. In any case, the integration with Google Analytics is nice; with one simple mouse click you can link your A/B test to an Analytics objective.
Automatic reporting
After starting your experiment, Google automatically dishes out variants of your website to visitors. You don't have to do anything for it and can now sit back - or spend your time on other marketing activities. When Google Optimize sees that a representative number of visitors have completed the test, you will receive a report in your mailbox.
Outsource your A/B testing
This blog post is a brief introduction to Google Optimize. As said, there are many possibilities and you can totally let your inner nerd run wild on this. If you don't have the technical knowledge for this or if you lack the time, then your A/B tests are in good hands at 2manydots.
At any given time, we run multiple A/B tests for our clients, which are essential to the optimization journey of delivered websites. Want to know more?Contact us for a 20-minute call.
