Let me start by saying that I’m a huge fan of Wix. They make it super easy to create websites. Even easier than WordPress!
This is because they use a web design model that was revolutionary for its time.
You know what? A/B testing, as a practice, is also going through the same change.
The traditional way to A/B test can be quite challenging for someone without any knowledge of programming and design, without mentioning that it takes a lot of time and resources to make it right.
That’s why there is one tool that hacked the A/B testing** process to make it super approachable, thanks to the use of AI.
It is the game-changer for the industry, and now I’m going to show you how you can set up your first A/B test with it, without spending a penny, and in less than ten minutes.
What is A/B tesing?
In Web design, A/B testing is a type of experiment that can be used to find out which changes on your Wix landing page (like headlines, copies, images, colors…) make your audience interact more.
While this seems like a great way of improving your conversion rate (thus, your revenue), it is not all rosy in the garden.
As I said earlier, the main downside in traditional A/B testing is that it can take a lot of time and knowledge to really nail it.
That’s why ABtesting.ai (the tool we are using today) was invented. To make this process as easy as possible.
With it, you can literally have a test running in your Wix site in less than 10 minutes.
Let’s go step by step, so I can show you how to use this tool to its max potential.
Step 1: Enter URL
The first step is to put the URL of the landing page you want to test in this box here:
And click on Start A/B testing
This will start a wizard that will guide you through the process of setting up your first A/B test.
Now, a bot is going to the landing page you entered in search of all the main elements that will have the greatest impact on your conversion rate if tested. These are the Headline, the main Copy, the main CTA, and the Hero Image.
Step 2: Choose Variations
Now, it’s time to start defining all the elements you want to test.
You’ll see that, as I already said, the tool will fetch all the main elements on your landing page to test.
If you see there’s one missing, or you feel like testing something completely different, click on Fetch More Elements (top-right corner), to visually select any other element you want to test.
You’ll see that the selected elements will have some texts already:
Those are text suggestions created by GPT-3, which is the most powerful AI when it comes to understanding and creating human-like copy.
When your page is fetched, GPT-3 scans the content, understands it and phrases some text suggestions you can try.
We use it to spark your creativity and help you avoid getting stuck in copywriting, which, otherwise, could take a lot of time and thinking.
Once you finish creating and editing all the elements you want to test, you click on Next.
Step 3: Choose the Conversion Goal
In this step is where you say to the tool: “Hey, when this happens, count it as a conversion, ok?”.
In other tools, selecting the conversion event can be quite tricky, and even require some coding experience.
ABtesting.ai tries to automatically fetch the main conversion event connected to the Wix landing page you are testing, so you don’t have to worry about any of that.
There are also two other ways of defining a conversion event: Target URL and by Code.
These are only recommended if your objective is to measure a very specific action that has not been fetched by the tool.
In this case, I recommend using Target URL, which means that you say to the tool: “When people get to this page after visiting the landing page I’m testing, count it as a conversion.”
Step 4: Integrate
This step is the one that frightens people the most, because they see a bunch of code they don’t understand, and start to panic.
That’s why we have a Wix plugin for it.
Just follow the steps here, and you’ll have it running in no time.
Then, you go back to your experiment and click on Verify.
Having integrated the snippet will allow the tool to actually run the experiment and measure results.
Once verified, you will get to the Overview. This means the test is successfully running.
The tool has created five variations using the elements you choose to test before, and started running a test that will measure their performance against each other and the original design, to see which of them converts the most.
Just like that. No more human input is needed from now on.
Also, with ABtesting.ai you get rid of another important problem that traditional A/B testing faces: Losing conversions while testing thanks to an underperforming variant.
Traditional A/B testing only allows you to determine a fixed amount of traffic to each variant (if there are five, 20% of the traffic goes to each one of them).
This can cause some conversion loss if there are some of the variations that underperform the original. All they do is losing conversions that the original would achieve.
ABtesting.ai uses something called the Multi-armed Bandit approach To solve this issue.
It basically means that the AI is capable of quickly understanding which variations are working better than the rest, and will drive more traffic to them, to compensate for the underperformers.
This way, you won’t lose conversions while testing, and also your test may take less time to complete.
One important thing you need to know is that an A/B test will only end when it reaches statistical significance.
This means that, depending on the conversion rate your page is having, it’ll need a certain amount of traffic to be confident about the fact of defining a variation as a winner.
And, as I said, when a test reaches statistical significance, it ends.
With traditional A/B testing, this means that, if you want to do a new test, you’ll have to start all over again: defining what to test, copywriting, dividing traffic, measuring, etc.
Luckily, ABtesting.ai uses Evolutionary Algorithm to learn from the results of the first experiment and, based on that, start a new test automatically.
This test will have new, better-performing combinations of the existing elements, to test against the winner of the previous experiment
This way it will always end up finding the best combination possible, meaning you get constant optimization. That’s it.
That’s everything you need to know about how to set up your first A/B test on your Wix site