How to test a new headline using A/B Testing

Back

The headline is one of the first things that visitors see when they arrive on your homepage. If you want to know which version of your headline converts more, here are the steps that you must follow:

Step 1. Go to the ‘Dashboard’ – >  ‘A/B Testing’.

omniconvert a b testing

Step 2. From the next window choose ‘Create A/B Test’ and enter a name for your experiment.

omniconvert create a b test

Next, you can either add the exact URL of the web page you intend to make changes to into the ‘Enter the page URL you want to edit’ field or add its HTML source code to start testing (‘upload your own HTML code‘ = mainly used for web pages generated dynamically, like the shopping cart page).

experiment add url html code

Step 3. Create the variation.

To apply the changes you wish (in our case, changing the headline), you must go to the ‘Variations’ tab and add a new Variation ( ‘+ NEW VARIATION’ ).
Here you will find the WYSIWYG editor which will help you make changes to the original version (‘CONTROL‘) into the new Variation.

Make the changes in the editor as we did in our example below.

A. Click on your headline and from ‘Edit element’, choose ‘Content’.

a b test edit content

When you click on ‘Content’, a new window will appear.

Here you can partially change the already existing content or insert a new one:

a b test edit content text

After you click on ‘Save’ the variation will look like this:

a b test content final look

B. Next, we show you how you can change the color to make it more visible. From ‘Edit element’, choose ‘CSS style’.

a b test edit element css style

In the new window that appeared, you can modify certain aspects regarding the overall aspect of the copywriting such as font-size, font-color, text-alignment etc.:

a b test edit css style

After you click on ‘Save’, the variation will look like this:

a b test css style final look

Step 4. Choose your audience.

Now that you’ve applied the modifications to your new variation you must manage the pages included in the experiment using rules.
In our example, we’re targeting the home page so the conditions should be URL is’ – http://www.website-name.com.

And if you have paid campaigns that use extra parameters for Analytics, you must also add ‘URL starts with’ – http://www.website-name.com? (adding the ‘?’ symbol at the end).

audience tab url condition setup

After the links are entered, select the ‘Segment’ (Target) of visitors that will be included in the experiment.

experiment segment target

When you click the drop-down button, a new window will appear. Here, you have to ‘Select a segment’. From this window, you can choose an existing segment or create a new one. If you wish to see how your experiment affects the users faster, we recommend leaving ‘All visitors‘ selected.

create new select existing segment

Under advanced settings, you can Limit the number of tested visitors, select the Device type and set the experiment as exclusive.

segment advanced settings exclusive

Step 5. Allocate traffic and set up goals.

In this tab, you must select the percentage of visitors that will see the experiment. This percentage will be randomly extracted from the total amount of users that have passed the segmentation you decided to use. You can run the test on 100% of the visitors who are included in the segment or just a part of them (let’s say 50%).

Then, you have to make sure that you split the traffic evenly between Control and Variation. Omniconvert does it automatically for you. If you want to check it by yourself, proceed like in this image:  

experiment split traffic

You have 3 main goals set from the start. If you want to add another one just click on ‘Select or create a goal‘:

experiment select create goal

Step 6. Schedule the A/B testing experiment.

Choose: a) for how long you want it to be live, b) when to be triggered for your visitors, c) whether or not you want to receive an email when it has expired.

a b test schedule

Step 7. Publish.

Finally, click on ‘Publish‘. You can also save the experiment as a draft to work on it later.

a b test publish save draft

DON’T FORGET: After each time you publish or update an experiment, you have to wait 3-4 minutes until it will be visible on your website. If you want to see the changes live instantly, install the OmniTool extension that helps bypass the cache system which generates the delay. You can download OmniTool here.

Shopping Basket