# A/B ou N testing

### 1. Business Value: Decide based on data, not intuition

A/B testing allows you to remove any doubts about the effectiveness of a change by comparing two (or more) versions with similar audience samples.

* **Maximising conversions**: Validate which headline or button generates the most clicks or purchases.
* **Consent optimisation:** Identify the banner design that maximises opt-ins while remaining compliant.
* **Risk reduction**: Test a new feature on a small percentage of your traffic before rolling it out across the board.

### 2. Implementation methodology

#### Step 1 : PREPARE THE DESIGN OF THE BANNERS TO BE TESTED 

* This first step consists of defining the format and design of the banners to be tested in your A/B test campaign. 
* Go to the Privacy Banners module to create your banners.

<figure><img src="https://1613709692-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2F0zW0zvGDZaiNuIFweCtu%2Fuploads%2FiAZnUYvoIhqT75RCONH6%2Fimage.png?alt=media&#x26;token=d2c51c42-3349-4a50-bb41-73320564fdd5" alt="" width="181"><figcaption></figcaption></figure>

#### YOU DON'T KNOW HOW TO SET UP YOUR BANNER ?

* Don't panic, we've got a ready-to-use tutorial so all you have to do is follow it step by step.

{% embed url="<https://doc.commandersact.com/fr/fonctionnalites/consent-management/setup-guides>" %}

#### Step 2 : CREATE THE DISPLAY SCENARIO FOR YOUR A/B or N TEST&#x20;

* This step consists of telling your container which test scenario to run.&#x20;
* To do this, go to the menu Data management > Data specification > Web datalayer > Data storage&#x20;
* Then click on "add data storage".

<figure><img src="https://1613709692-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2F0zW0zvGDZaiNuIFweCtu%2Fuploads%2FwrZTFGtWtceVY7Sgyt3p%2Fimage.png?alt=media&#x26;token=4282b8f2-9bce-4054-b58e-896bfef7ba65" alt=""><figcaption></figcaption></figure>

* Select the scenario you wish to deploy&#x20;
  * Either, A/B Test: 50%/50% :&#x20;

<figure><img src="https://1613709692-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2F0zW0zvGDZaiNuIFweCtu%2Fuploads%2FeMpV8UEu1nPaPruN4P04%2Fimage.png?alt=media&#x26;token=fb1391f6-d8f4-40ff-8de0-a2225706318d" alt="" width="310"><figcaption></figcaption></figure>

* That is, N Test: A 80% / B 10% / C 10% :

<figure><img src="https://1613709692-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2F0zW0zvGDZaiNuIFweCtu%2Fuploads%2F0eNXaI7rYLeVvChI74kY%2Fimage.png?alt=media&#x26;token=c0338457-c6e4-4228-a566-a35323ec5204" alt="" width="318"><figcaption></figcaption></figure>

***

#### Step 3 : ADD PRIVACY DISPLAY CONSTRAINTS ON COOKIES&#x20;

* The purpose of this step is to define the display scenario for the banners you created [in step 1](#step-1).&#x20;
* To do this, simply go to your container from the source > overview menu

<figure><img src="https://1613709692-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2F0zW0zvGDZaiNuIFweCtu%2Fuploads%2FpD2DbcT7LhyhtrUUGrCM%2Fimage.png?alt=media&#x26;token=6a34b627-0d6b-4784-9e11-0810771a70a2" alt="" width="375"><figcaption></figcaption></figure>

* Then define your display constraint, as shown in the example below, from the Rules > Privacy Constraint menu.

<figure><img src="https://1613709692-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2F0zW0zvGDZaiNuIFweCtu%2Fuploads%2Flg42Fqguisr6szUiSwok%2Fimage.png?alt=media&#x26;token=a688999d-94e3-4c47-a87f-d936c5489674" alt=""><figcaption></figcaption></figure>

* Choose the name for step 2 on the Name cookie :&#x20;

<figure><img src="https://1613709692-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2F0zW0zvGDZaiNuIFweCtu%2Fuploads%2FtqRm9BUA2taOo7jcrMFz%2Fimage.png?alt=media&#x26;token=716b836d-1570-411b-99fd-6b4c00aaf56b" alt=""><figcaption></figcaption></figure>

***

#### Step 4 : ACTIVATE YOUR BANNERS TO BE TAKEN INTO ACCOUNT IN YOUR SCENARIO

* In the Generation section, activate the. Banners for A/B Test

<figure><img src="https://1613709692-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2F0zW0zvGDZaiNuIFweCtu%2Fuploads%2FWSRiRMI8zED5N2ka3ZUL%2Fimage.png?alt=media&#x26;token=7c95eb76-4422-400f-932e-2950c0f11d41" alt=""><figcaption></figcaption></figure>

#### Step 5 : DEPLOY YOUR CONTAINER TO PUT YOUR TEST SCENARIO INTO PRODUCTION&#x20;

* In the Generation section, activate the A/B Test banners by deploying your container on your website!

<figure><img src="https://1613709692-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2F0zW0zvGDZaiNuIFweCtu%2Fuploads%2F3O4E9QXfypnIMO20VsA0%2Fimage.png?alt=media&#x26;token=f797522c-1480-4797-a723-1a054b98ca29" alt=""><figcaption></figcaption></figure>

#### Need help setting up your first test?

Our consultants will help you define your test protocols and interpret statistical results to ensure reliable decisions. **Contact your dedicated consultant or our support team:** <support@commandersact.com>