Guide: Creating a Split Test

AB Tasty offers several types of testing campaigns designed to evaluate different aspects of your website or different hypotheses about what drives engagement and conversions. 

A/B testing is a method of testing any aspect of a website to find out if an alternative approach will increase engagement and conversions. A/B testing is a great way to find out which improvements are worth investing in or to determine what is driving engagement and conversions on a website. 

If you’re not sure which type of test to choose, our virtual assistant Ally can help you decide.

A split test is one of the A/B test campaigns types you can create within the AB tasty platform 

A Split Test consists of redirecting traffic from one or several pages of your website to new versions of these pages that you want to test. Visitor activity on the new pages is measured to evaluate whether they are more or less effective than the original versions of these pages. This type of campaign is similar to a classic A/B test but instead of creating a new variation using AB Tasty’s visual or code editor, the new pages need to be developed on the client side and hosted on your own servers.

Part 1: The main ways to create a new campaign

Split Tests, just like other types of campaigns, can be created directly from the campaign dashboard, audience manager dashboard, Chrome extension, or by duplicating or transforming an existing campaign. 

Directly from the dashboard

On the Test dashboard, hover over the Create button to bring up the following list. 

Click A/B Test to enter the A/B test creation workflow. 

📎Note: A Split Test is a type of A/B test. To create a Split Test, you will set up an A/B test and activate the Split Test option in the visual editor during this setup. 

Using the Chrome extension

AB Tasty has released a Chrome extension that allows you to create a campaign directly from the specific page of your website. 

To download the Chrome extension, please follow this link. To get more information about the AB Tasty Chrome extension, please visit our article. This option is useful to load logged pages.

Heads up ⚡️

To create a Split Test, you need to activate an option in the visual editor. Don’t worry if your page doesn’t load properly in the editor as this is not a problem.

Duplications and transformations

If you want to duplicate a former A/B test to create a new one (with or without modifying some elements of the set-up), select the Duplicate option on your dashboard. 

• Click the three dots on the right of the campaign line you want to duplicate.

• Click on the first option – Duplicate.
  
   

• The duplication modal opens. You can modify the name of your new campaign, otherwise, the name will be the same as the existing campaign but with (duplicate) added to the end.

• Select the account where you wish to store your new campaign.
  
  

For more information about duplications and transformation, refer to our specific article on this topic.

Part 2: The Main Information Page

The Main Information Page is the first step in setting up your new A/B test campaign. 

General use case: Pasting in the URLs you wish to test

At the top of the page, the dropdown is set to A/B test. At this stage, you can select another kind of test if you’ve decided to run a different type of test. 

The fields Name and Sample URL are mandatory and need to be filled in to access the Save button. 

The Sample URL field should be filled in with the URL of the exact page you want to test or a sample of the pages that share the same layout as the page you want to test. It will be used in your campaign: 

• To load the visual editor.

• As a pre-filled targeted URL in the targeting step, but it won’t be the final targeted page of your campaign. See our article on targeting for more information. 

❗Caution: Don’t use URL parameters in the sample URL fields, otherwise you won't be able to save the step.

⭐Tip: Filling the Hypothesis field is a best practice to help collaboration between teams so everyone can easily understand the objective of your A/B test (for example, testing a new layout for your product pages). This will avoid wasting time looking for the changes in the visual editor.

Good to know 💡

Once you click on Save, your A/B test will be generated (with a unique ID) and will appear in your Tests dashboard. If you return to the Main Information Page, everything will be editable except the type of test (this dropdown won’t be available anymore). At this stage, if you need to change the type of campaign, you should restart the campaign creation process from scratch or transform your current A/B test campaign to another kind of test using the duplication action.

Part 3: The content of your campaign

To run a Split test, you don’t need to create any specific content. The objective of this setup is to use the classic A/B test campaign workflow to activate a redirection of traffic from the former page (the original) to the new one(s) (the variations).

Using the Visual Editor:

To learn how to activate a redirection (variation B of your A/B test) or several redirections (variations C, D, E, F, etc.), refer to the section on Redirect options in our Visual Editor Guide. 

Part 4: Goals

Once you’ve created your redirections in the visual editor, you can pre-configure your future reporting to easily identify the metrics needed to track your primary goal (that which you are hoping to impact most with your A/B test). For example: if your Split Test consists of testing a new landing page layout, your primary goal should be the bounce rate, along with optional secondary goals. 

To learn about different types of metrics and how to set them up, see the following guide.

Metrics are categorized into different sections: 

• Action tracking: All the trackers you’ve set up in the visual editor – note, these are not retroactive.

• Page tracking: The pageviews you’ve recorded in your account plus the option to create new ones (for more on this, see our guide). These are retroactive.

• Custom tracking: The action tracking and any other tracking you’ve set up by adding code in the editor or the global code of the account (for more on this, see our guide). These are not retroactive.

• Browsing metrics, including bounce rate, pages per session, and visited rate. To learn more about these metrics and how they are calculated, read the following guide. These are retroactive.

• Transaction metrics: Here you’ll find the transaction indicators you’ve set up in your account. To use these, you need to install our transaction tag on your website. For more on this, see the following guide. Note, these are also retroactive. 

Good to know💡

As the URL of the original version of your page is different from its variations, the code will probably be different. If you want to track clicks on the main CTA of both pages, follow these steps: 

• First, add action tracking to the original page. You can access the original page in the editor – it’s displayed by default in the Original tab in the top banner of the visual editor. As with every tracker, AB Tasty will try to detect the presence of the tracked element in both the original and the variation pages, even if this element is present on only one of the two.
• Next, do the same for the variation you want to test. If you return to the Variation tab, you will see the original version of your page – that’s normal; it’s because you’ve used this URL in the Main Information Page. To load the new one, go back to the Main Information Page and paste the new URL in the field named URL to load in the editor, and click save. Then go back to the editor. Your new page will be loaded in the Variation tab. 
• Now you can add your tracker in the variation in the usual way. Use the same name as the tracker in the original version. This way all data collected on the original element and the new element will contribute to the same metric, even though they are hosted on different pages, with different IDs.
• Another solution is to code your trackers directly by adding JavaScript (global code added in the right side panel).

Tips ⭐

For the most efficient workflow in setting up an A/B test, list all your metrics and their trackers on a separate sheet or document (for example metric “click rate”, so tracker “click tracking on the element #mainCTA). Then implement all the trackers in the original variation, before finally implementing them all in the variation. 

Each time, you can copy and paste the tracker names from the list, to ensure none are overlooked and your associated metrics will have the same names. This is necessary to ensure all relevant data is collected under the same KPI in your campaign report.

Selecting your primary and secondary goals

During this step, you will need to select a primary goal, which will serve as a point of reference for your campaign. You can also select several secondary goals that can be tracked in the reporting of your campaign. You can select just the goals that interest you, and only they will appear in your campaign report. 

You can add or delete secondary goals later, after the launch of the campaign (in this case, they will be retroactive), but the main goal cannot be changed once it has been saved. However, the goals set up in the editor (Click tracking, Action Tracking) are not retroactive so you have to configure them before launching your test.

• Click the category related to your goal: A list of indicators appears on the right-hand side.

• Click the + icon to the right of the indicator you want to select. The goal is displayed in the right-hand pane as goal number 1 in the Primary goal category.
  
  

❗Caution: Once your goals have been saved and your campaign has started, you can no longer modify or delete the primary goal of your campaign. 

Once your campaign has launched, you can add secondary goals, but only pageviews will be retroactive in the reports. All other goals will be calculated from the date you’ve added these metrics as goals. You can add up to 49 secondary goals.

Save before leaving the page. Your set-up will constitute the basis of your future reporting to analyze the result of your campaign. Each chosen metric will appear as a goal in the reporting.

Hiding transaction indicators

When you implement a transaction tag, the transaction affiliation (TA) name you assigned to your transaction tag automatically appears as a transaction indicator in the Transaction section of the Goals setup step of a campaign.

For more information on how to configure a transaction tag, see our article on Installing a transaction tag.

The more transaction tags you implement, the longer the transaction indicators list will be. These transaction indicators can then be chosen as primary or secondary goals for your campaign.

The transaction indicators hide option enables you to clear and lighten the transaction indicators list by hiding out-of-date transaction indicators.

This makes it easier to find the transaction indicator(s) you want to set as primary or secondary goals for your campaign and to get rid of any obsolete indicators.

Before deleting transaction indicators, and if you have made any modification in the step (such as adding or removing a goal), click Save to ensure all goals are up-to-date. To hide transaction indicators from the Transaction list, follow these steps:

• Click Hide transaction indicators

• From the pop-in, hover over the indicator you want to hide from the list and click the trash icon

• Repeat this action for the other indicators you want to delete

• Click the Hide button to confirm

Heads up⚡

You can’t hide an indicator that is used in one or several other campaigns, whether they are live or paused, therefore the option to hide these is disabled. 

Good to know💡

Hidden transaction indicators will be hidden from this list for all campaigns but not deleted permanently. If you wish to delete one permanently, you should do it directly in your website code.

Retrieving hidden transaction indicators

You can retrieve the list of all transaction indicators you have hidden in your account settings by visiting Account management > Transaction indicators management.

To reactivate a hidden transaction indicator, hover over it and click Activate. The indicator will disappear from this page and you’ll be able to see it in the transaction indicators list in the Goals setup step of a campaign. You will now be able to use it as a primary or secondary goal for your campaigns.

Part 5: Targeting

Targeting is one of the most important moments of the campaign setup. This determines WHERE your campaign is visible on your website (on every page, on specific pages, on only one page, etc.), for WHOM it is visible(all traffic, mobile users only, etc.), and in WHICH conditions the campaign is displayed (after a certain number of pages, when the weather outside drops below a certain temperature, etc.).

Please refer to our complete guide about targeting to learn how to set up segments of visitors, saved pages, and saved triggers to save time in future campaign setups.

Good to know💡

• The segment and the trigger that you choose in the WHO and HOW sections work the same way as in other types of campaigns.
• For targeted pages, you must input the URL(s) of the original pages that you want to redirect to your variations. 
• For example, let’s say you want to test a new layout for your product pages.
• The current URL is www.mywebsite.com/product-XXXXX (where XXXXX is the ID of the product). 
• Your new URL is: www.mywebsite.com/product2-XXXXX
• In this case, you will input the new URL in the visual editor (see the following section) and the targeted URL will be: regex matches www.mywebsite.com/product/[0-9]*

Part 6: Traffic allocation

The traffic allocation step enables you to choose if you want to personalize all your targeted traffic or just a small part of it. It also allows you to change the ISO repartition between variations if you need to.

To learn more about traffic allocation, read our dedicated guide. 

Part 7: Advanced options

📎Note: Advanced options are not mandatory. 

Third-party integrations (analytics)

You can link your Split Test to your analytics tool, allowing you to analyze your cohorts in your analytics dashboards. AB Tasty can easily be integrated with many popular analytics tools. The API connection can be set up at the account level, which will apply the integration to all subsequent campaigns automatically. To learn more, refer to the following guide.

If you want to set up analytics integration for only one specific A/B test, find your analytics tool in the dropdown and select it. Some analytics tools, such as Google Analytics, require additional settings, which will appear in a form panel:

Once launched, your test will send data to all the analytics services that you have configured here.

Part 8: QA

The QA of a campaign is one of the most important steps to take during the campaign set-up. To learn everything about this step, please refer to our complete guide on QA & QA Assistant.

Part 9: Scheduler

If you’d like your Split Test to start at a specific time on a specific day, or for it to pause at a specific time, see our guide on scheduling.