Set up an A/B Test in a web or inpage deployment

To optimize your response rate, the platform offers you the possibility to set up an A/B Test for your web or inpage deployment in just a few clicks. This is to determine the ideal formula for your needs.

We understand that finding the right combination of survey configurations and triggering conditions can be quite challenging. To help you discover the winning formula, we suggest conducting A/B testing by setting up your survey deployments with random triggering conditions and managing visitor solicitation.

This random feature allows you to evenly distribute your visitors into populations in order to :

  • test multiple survey formats simultaneously (short, long, etc.)
  • test multiple types of deployments simultaneously (pop-in vs button, etc.)

Requirements

  • Have completed the construction of your survey(s) and finished configuring your campaign(s)
  • Have an up-to-date deployment tag and ensure that your web deployment display correctly overall on your site

Default management of randomness

The configuration of randomness on the platform follows a certain logic, which, once understood, allows for simple and intuitive setting of your random trigger conditions.

The order of configuration of random elements is very important because the oldest condition always takes priority and will be executed first.

This random feature of the platform is based on non-cumulative (disjoint) probability management. Therefore, the number of elements will define the probabilities to be configured to manage the random trigger :

Probability Setup
1 in 2
  • Element A at 1/2 or 50%
  • Element B without condition or at 100%
1 in 3
  • Element A at 1/3 or 34%
  • Element B at 1/2 or 50%
  • Element C without condition or at 100%
1 in 4
  • Element A at 1/4 or 25%
  • Element B at 1/3 or 34%
  • Element C at 1/2 or 50%
  • Element D without condition or at 100%
1 in 5
  • Element A at 1/5 or 20%
  • Element B at 1/4 or 25%
  • Element C at 1/3 or 34%
  • Element D at 1/2 or 50%
  • Element E without condition or at 100%
1 in 6
  • Element A at 1/6 or 17%
  • Element B at 1/5 or 20%
  • Element C at 1/4 or 25%
  • Element D at 1/3 or 34%
  • Element E at 1/2 or 50%
  • Element F without condition or at 100%
1 in 7
  • Element A at 1/7 or 15%
  • Element B at 1/6 or 17%
  • Element C at 1/5 or 20%
  • Element D at 1/4 or 25%
  • Element E at 1/3 or 34%
  • Element F at 1/2 or 50%
  • Element G without condition or at 100%
1 in 8
  • Element A at 1/8 or 13%
  • Element B at 1/7 or 15%
  • Element C at 1/6 or 17%
  • Element D at 1/5 or 20%
  • Element E at 1/4 or 25%
  • Element F at 1/3 or 34%
  • Element G at 1/2 or 50%
  • Element H without condition or at 100%
1 in 9
  • Element A at 1/9 or
  • Element A at 1/9 or 12%
  • Element B at 1/8 or 13%
  • Element C at 1/7 or 15%
  • Element D at 1/6 or 17%
  • Element E at 1/5 or 20%
  • Element F at 1/4 or 25%
  • Element G at 1/3 or 34%
  • Element H at 1/2 or 50%
  • Element I without condition or at 100%
1 in 10
  • Element A at 1/10 or 10%
  • Element B at 1/9 or 12%
  • Element C at 1/8 or 13%
  • Element D at 1/7 or 15%
  • Element E at 1/6 or 17%
  • Element F at 1/5 or 20%
  • Element G at 1/4 or 25%
  • Element H at 1/3 or 34%
  • Element I at 1/2 or 50%
  • Element J without condition or at 100%

etc.

Conduct an A/B Test to distribute half of the visitors across 2 web or inpage deployments

The configuration of randomness on the platform follows a certain logic, which, once understood, allows for simple and intuitive setting of your random trigger conditions.

The order of configuration of random elements is very important because the oldest condition always takes priority and will be executed first.

Thus, to configure an A/B test distributing half of the visitors evenly between survey A and survey B, follow the procedure described below.

Step 1 : Create Deployment A at 50%

First, start by creating the first deployment (Web Deployment A) of your campaign.

In the "IF" condition, press the "Add new condition" button.

Then choose the "Random" condition.

Leave the "Mode" field on "Visit" and select a probability of "1 for 2" if expressed in fraction or a probability of "50%" if expressed in percentage.

In the "THEN" condition, add a new condition by selecting the display type of your survey on the website or the inpage display settings.

Once the configurations of your display mode are finalized, your first deployment (Web Deployment A) is now ready.

Step 2 : Create Deployment B by managing solicitation

Create a second deployment (Web Deployment B) of your campaign without a random trigger condition but with a solicitation condition on the campaign in question.

This will automatically display Deployment B if Deployment A does not run.

To do this, in the "IF" condition of your Web Deployment B, press the "Add new condition" button.

Then choose the "Solicitation" condition.

Leave the "Visitor" field on "has already been". Depending on whether your A/B test is conducted between two different campaigns or only two different deployment types, select the appropriate element in the "solicited by" field :

Then choose the campaign or deployment on which the display of your Web Deployment B will depend and complete the time range. We chose "360 days" because it is the default cookie retention period.

Adapt the solicitation condition based on the behavior of visitors on the page where the survey is displayed.

If the page is regularly visited by the same user, use "has already been" : a visitor who has already been solicited by Deployment A will automatically see Deployment B. On the other hand, if the page is rarely revisited, prefer "has never been" : visitors who were not included in the random condition of Deployment A will automatically see Deployment B.

In the "THEN" condition, add a new condition by selecting the display type of your survey on the website or the inpage display settings.

Once the configurations of your display mode are finalized, your second deployment (Web Deployment B) is now ready.

 

Your A/B test configuration is now complete, you just need to activate your 2 deployments and monitor their results to evaluate which one is the most effective.

Conduct an A/B Test to distribute a third of the visitors across 3 web or inpage deployments

The configuration of randomness on the platform follows a certain logic, which, once understood, allows for simple and intuitive setting of your random trigger conditions.

The order of configuration of random elements is very important because the oldest condition always takes priority and will be executed first.

Thus, to configure an A/B test distributing a third of the visitors evenly between survey A, survey B, and survey C, follow the procedure described below.

Step 1 : Create Deployment A at 34%

First, start by creating the first deployment (Web Deployment A) of your campaign.

In the "IF" condition, press the "Add new condition" button.

Then choose the "Random" condition.

Leave the "Mode" field on "Visit" and select a probability of "1 for 3" if expressed in fraction or a probability of "34%" if expressed in percentage.

In the "THEN" condition, then add a new condition by selecting the display mode of your survey on the website or the inpage settings.

 

Once the configurations of your display mode are finalized, your first deployment (Web Deployment A) is now ready.

Step 2 : Create Deployment B at 50% by managing solicitation

Here, the goal is to target half of the remaining 2/3 of visitors not solicited by Web Deployment A.

To do this, create a second deployment (Web Deployment B) of your campaign.

In the "IF" condition, press the "Add new condition" button.

Then choose the "Random" condition.

Leave the "Mode" field on "Visit" and select a probability of "1 for 2" if expressed in fraction or a probability of "50%" if expressed in percentage.

Then, below your random condition in the "IF" condition of your Web Deployment B, press the "Add new condition" button.

Then choose the "Solicitation" condition.

Leave the "Visitor" field on "has already been". Depending on whether your A/B test is conducted between three different campaigns or only three different deployment modes, select the appropriate element in the "solicited by" field :

Then choose the campaign or deployment on which the display of your Web Deployment B will depend and complete the time range. We chose "360 days" because it is the default cookie retention period.

Adapt the solicitation condition based on the behavior of visitors on the page where the survey is displayed. If the page is regularly visited by the same user, use "has already been" : a visitor who has already been solicited by Deployment A or B will automatically see Deployment C. On the other hand, if the page is rarely revisited, prefer "has never been" : visitors who were not included in the random condition of Deployment A or B will automatically see Deployment C.

The operator between the two "IF" condition must be set to "AND".

In the "THEN" condition, then add a new condition by selecting the display mode of your survey on the website or the inpage settings.

 

Once the configurations of your display mode are finalized, your second deployment (Web Deployment B) is now ready.

Step 3 : Create Deployment C by managing solicitation

Finally, create a third deployment (Web Deployment C) of your campaign without a random trigger condition but with a solicitation condition on the campaign in question.

This will automatically display this deployment C if deployments A and B do not execute.

To do this, in the "IF" condition of your Web Deployment C, press the "Add new condition" button.

Then choose the "Solicitation" condition.

Leave the "Visitor" field on "has already been". Depending on whether your A/B test is conducted between three different campaigns or only three different deployment modes, select the appropriate element in the "solicited by" field :

Then choose the campaigns or deployments on which the display of your Web Deployment C will depend and complete the time range. We chose "360 days" because it is the default cookie retention period.

Adapt the solicitation condition based on the behavior of visitors on the page where the survey is displayed. If the page is regularly visited by the same user, use "has already been" : a visitor who has already been solicited by Deployment A or B will automatically see Deployment C. On the other hand, if the page is rarely revisited, prefer "has never been" : visitors who were not included in the random condition of Deployment A or B will automatically see Deployment C.

In the "THEN" condition, then add a new condition by selecting the display mode of your survey on the website or the inpage settings.

 

Once the configurations of your display mode are finalized, your third deployment (Web Deployment C) is now ready.

 

Your A/B test configuration is now complete, you just need to activate your 3 deployments and monitor their results to evaluate which one is the most effective.

Conduct an A/B Test to distribute a quarter of the visitors across 4 web or inpage deployments

Thus, to configure an A/B test distributing a quarter of the visitors evenly between survey A, survey B, survey C, and survey D, follow the procedure described below.

Step 1 : Create Deployment A at 25%

First, start by creating the first deployment (Web Deployment A) of your campaign.

In the "IF" condition, press the "Add new condition" button.

Then choose the "Random" condition.

Leave the "Mode" field on "Visit" and select a probability of "1 for 4" if expressed in fraction or a probability of "25%" if expressed in percentage.

In the "THEN" condition, then add a new condition by selecting the display mode of your survey on the website or the inpage settings.

 

Once the configurations of your display mode are finalized, your first deployment (Web Deployment A) is now ready.

Step 2 : Create Deployment B at 34% by managing solicitation

Here, the goal is to target half of the remaining 3/4 of visitors not solicited by Web Deployment A.

To do this, create a second deployment (Web Deployment B) of your campaign.

In the "IF" condition, press the "Add new condition" button.

Then choose the "Random" condition.

Leave the "Mode" field on "Visit" and select a probability of "1 for 3" if expressed in fraction or a probability of "34%" if expressed in percentage.

Then, below your random condition in the "IF" condition of your Web Deployment B, press the "Add new condition" button.

Then choose the "Solicitation" condition.

Leave the "Visitor" field on "has already been". Depending on whether your A/B test is conducted between four different campaigns or only four different deployment modes, select the appropriate element in the "solicited by" field :

Then choose the campaign or deployment on which the display of your Web Deployment B will depend and complete the time range. We chose "360 days" because it is the default cookie retention period.

Adapt the solicitation condition based on the behavior of visitors on the page where the survey is displayed.

If the page is regularly visited by the same user, use "has already been" : a visitor who has already been solicited by Deployment A will automatically see Deployment B. On the other hand, if the page is rarely revisited, prefer "has never been" : visitors who were not included in the random condition of Deployment A will automatically see Deployment B.

The operator between the two "IF" condition must be set to "AND".

In the "THEN" condition, then add a new condition by selecting the display mode of your survey on the website or the inpage settings.

 

Once the configurations of your display mode are finalized, your second deployment (Web Deployment B) is now ready.

Step 3 : Create Deployment C at 50% by managing solicitation

Here, the goal is to target half of the remaining 2/4 of visitors not solicited by Web Deployments A and B.

To do this, create a third deployment (Web Deployment C) of your campaign.

In the "IF" condition, press the "Add new condition" button.

Then choose the "Random" condition.

Leave the "Mode" field on "Visit" and select a probability of "1 for 2" if expressed in fraction or a probability of "50%" if expressed in percentage.

Then, below your random condition in the "IF" condition of your Web Deployment C, press the "Add new condition" button.

Then choose the "Solicitation" condition.

Leave the "Visitor" field on "has already been". Depending on whether your A/B test is conducted between four different campaigns or only four different deployment modes, select the appropriate element in the "solicited by" field :

Then choose the campaign or deployment on which the display of your Web Deployment C will depend and complete the time range. We chose "360 days" because it is the default cookie retention period.

Adapt the solicitation condition based on the behavior of visitors on the page where the survey is displayed.

If the page is regularly visited by the same user, use "has already been" : a visitor who has already been solicited by Deployment A or B will automatically see Deployment C. On the other hand, if the page is rarely revisited, prefer "has never been" : visitors who were not included in the random condition of Deployment A or B will automatically see Deployment C.

The operator between the two "IF" conditions must be set to "AND".

In the "THEN" condition, then add a new condition by selecting the display mode of your survey on the website or the inpage settings.

 

Once the configurations of your display mode are finalized, your third deployment (Web Deployment C) is now ready.

Step 4 : Create Deployment D by managing solicitation

Finally, create a fourth deployment (Web Deployment D) of your campaign without a random trigger condition but with a solicitation condition on the campaign in question.

This will automatically display this deployment D if deployments A, B, and C do not execute.

To do this, in the "IF" condition of your Web Deployment D, press the "Add new condition" button.

Then choose the "Solicitation" condition.

Leave the "Visitor" field on "has already been". Depending on whether your A/B test is conducted between three different campaigns or only three different deployment modes, select the appropriate element in the "solicited by" field :

Then choose the campaigns or deployments on which the display of your Web Deployment D will depend and complete the time range. We chose "360 days" because it is the default cookie retention period.

Adapt the solicitation condition based on the behavior of visitors on the page where the survey is displayed.

If the page is regularly visited by the same user, use "has already been" : a visitor who has already been solicited by Deployment A, B or C will automatically see Deployment D. On the other hand, if the page is rarely revisited, prefer "has never been" : visitors who were not included in the random condition of Deployment A, B or C will automatically see Deployment D.

In the "THEN" condition, then add a new condition by selecting the display mode of your survey on the website or the inpage settings.

 

Once the configurations of your display mode are finalized, your fourth deployment (Web Deployment D) is now ready.

 

Your A/B test configuration is now complete, you just need to activate your 4 deployments and monitor their results to evaluate which one is the most effective.

 

If you have any questions, feel free to contact our Customer Care team.

 

Was this article helpful?
0 out of 0 found this helpful