A personalization is a type of campaign which enables you to modify the content of your website for all of your traffic or for a specific audience.
Traffic allocation is managed differently for personalizations than for A/B tests. For A/B tests, users are randomly allocated to the original version or to a variation depending on the defined targeting rules.
For personalizations, 100% of traffic is allocated to the personalization depending on the defined targeting rules.
Personalizations can be created based on an A/B test for two main reasons:
- If you’ve established that the test results identify a winning variation for a specific audience.
For example, if your A/B test is neutral: this means there is no significant improvement of any variation compared to the original version. However, it shows significant improvement for a given audience (for instance with mobile users only). In this case, you can transform the variation into a personalization and target this particular audience.
- If you’ve established that the test results identify a winning variation: you can transform the winning variation into a “stopgap” personalization in order for 100% of your traffic to be subjected to the best version while the technical development team works on implementing the change.
“Stopgap” personalization is a temporary solution. We do not recommend using this personalization on a long term basis.
Moreover, when you identify one of the variations of your A/B test as performing better than the others, rather than reallocating 100% of the traffic to it, we recommend converting it into a personalization for the following reasons:
- You would lose the data of your initial test and would struggle to extract it if you needed to justify your decision.
- Users allocated to the original version or to another variation would never see the winning version because the cookie guarantees permanent exposure to one of the versions from the very first session. In this case, only new visitors or visitors already allocated to the winning variation from the start will be targeted.
This is why the best solution is to create a new personalization campaign based on your winning variation, yet unrelated to the test.
Transforming a variation into a personalization
To create a personalization based on the variation of an A/B test, apply the following steps:
- From the dashboard, click the (Edit) icon of the test you want to transform into a personalization.
- Click Variation editor
- In the variation editor, click the variation you want to use.
- In the variation menu, select the Create personalization option.
- Click Yes to confirm.
The personalization is created and appears in the Personalization section of your dashboard. It keeps the name of the A/B test it was created based on.
The personalization is always named as a function of the A/B test and not as a function of the chosen variation.
However, the changes applied to the variation you have chosen to transform into a personalization are taken into account.
Managing personalization options
Configuring the scenario
The variation editor opens automatically when you convert the variation into a personalization. To configure the personalization, click the scenario tab (i.e. your former A/B test variation) and use the drop-down menu if necessary.
The variation editor enables you to configure changes in the same way as a standard A/B test (refer to Features and variation modifications index).
To access targeting, click Targeting (step 3).
This step enables you to define an audience in order for changes to be visible to a specific visitor segment only. This audience may be defined as a function of behaviors identified in the reports of the A/B test the personalization is based on.
For more on this, refer to Defining your audience.