# A/B testing
### What is the A/B testing?
The A/B testing is a **powerful tool** that aims to **improve both the open rate and the click rate** of your campaigns.
Briefly explained, you create **two versions of a campaign**, and AcyMailing will send them to a sample of your recipients, wait for some statistics to be collected, then it will **send the best version** to the rest of the recipients.
### How to set it up
You'll first need to create a campaign like you are used to, but select the "A/B testing" campaign type:
Once you're happy with your first version, you can begin creating the second version of your campaign by clicking "Version B":
The second version can have a different subject, mailbox listing preview line and a different content. You can copy the version A to make subtle changes, or even start from an other template:
### Send settings
Once both versions are ready, select the lists you want your campaign to be sent to, then on the "Send settings" step you will have to choose how AcyMailing will handle your campaign.
The first option lets you choose the size of the sample that will receive the test versions. In this example, the campaign is sent to 10000 users and we chose to use 10% of them as a sample.
10% of 10000 is 1000, so **500 recipients will receive the version A, and 500 will receive the version B**.
AcyMailing will then wait for two days then compare the statistics of both versions of your campaign.
In this example, we choose to send the best version: we will take the subject of the version with the higher open rate, and the content of the version with the higher click rate.
Here is the summary we have with this example:
### What do I need to do next?
Once you confirmed the A/B testing campaign, there is nothing more to do as AcyMailing will take care of the rest. After the delay you chose for the first statistics to be collected, a new campaign will be generated with the "winning version".
You can find your A/B campaigns on the campaigns listing, they will have a vial next to them:
