> ## Knowledge Base Index
> Fetch the complete knowledge base index at: https://help.avada.io/sitemap.xml
> Use this file to discover available pages before exploring further.
> Pure-Markdown content can be obtained by appending a '.md' suffix to the content URLs listed in the sitemap (without the trailing slash).

# How to set up A/B Testing Split in Automation workflows?

A/B testing split allows you to find out what combination is best for your workflow. You can set up the percentage for each path then customer data will flow down to the paths based on percentage. Let's see how we can set up it in AVADA.

# 1 Add A/B testing to the workflow

## 1.1 New Automation

**Step 1:** **App > Automation > New workflow** > **Select the preset** > Click **Next**


![](https://storage.crisp.chat/users/helpdesk/website/b66870a37c890800/1_1omlufv.png)

**Step 2:** Drag and drop **A/B testing** into the workflow


![](https://storage.crisp.chat/users/helpdesk/website/b66870a37c890800/2_888bek.gif)

**Step 3:** Edit split


![](https://storage.crisp.chat/users/helpdesk/website/b66870a37c890800/3_1ngc55.png)

-   **A/B Testing Name**: Enter a name for identify your A/B testing
-   **Edit the scale of paths**: Drag the sliders on paths A and path B to edit the ratio or you can enter the percentage you want

-   You can delete a path ( path A, path B) or both paths. Please click on **Delete** > Select the path you want to delete or select both

![](https://storage.crisp.chat/users/helpdesk/website/b66870a37c890800/fc0wm7s_11z12qo.png)

-   Click **Save** to save the edited result

**Step 4**: Click **Next** and continue to complete the campaign

## 1.2 Old Campaigns

If you want to add **A/B Testing** to the old campaign:
**Step 1:** **App > Automation > Select Workflow Tab **
**Step 2**: Follow steps 2 and 3 as above
**Step 3:** Then click **Save**

# Activities of A/B testing


![](https://storage.crisp.chat/users/helpdesk/website/b66870a37c890800/4_1xhgr9b.png)

Let see the automation workflow above and let us explain how A/B testing works:

When a customer creates a new order on your store, 50% of customers will receive an email on path A, 50% of customers will receive an email on path B, but this is completely random distribution. You can see the results of each email/sms right under them to easily compare and find out the best ones.