Testing the zone placement

Tests how visitors respond to recommendations that are placed in different zones.

About this task

You create two zones. 50% of the time recommendations are displayed in a single zone and 50% of the time recommendations are displayed in the other zone. Visitors see the same zone placement during the length of the test. For example, test if the horizontal zone is more effective than the zone at the bottom of the page.

Procedure

  1. This test assumes that both zones are coded into the site's pages and are tested to be fully functional with the ability to render Digital Recommendations dynamic recommendations.
  2. This test requires the creation of two A/B tests. Create A/B Test 1 according to the following steps. You must assign No Recommendations to the opposite test group in Test 1 versus Test 2. The percentage splits must be the same for the control and test groups across both tests. This split ensures that any visitor who falls into the Control of Test 1 also fall into the Control of Test 2. Likewise, any visitor who falls into Test A group of Test 1 also fall into Test A group of Test 2.
    1. Create the first test around zone 1 where the Control Group sees the default set of recommendations 50% of the time.
    2. Configure Test A to see No Recommendations the other 50% of the time. The "No Recommendations" option is available in the Recommendation Plan list in the A/B Test Settings. When "No Recommendations" is selected, Digital Recommendations delivers "_NR_" to the site's zone population function. This response causes the site zone not to display.
  3. Create A/B Test 2 as follows:
    1. Create a second A/B test around zone 2 where the Control shows No Recommendations 50% of the time.
    2. Configure Test A to use the recommendation plan that is used on the Control Group of Test 1.