What is A/B Testing ?
An « A/B Test », or called an « A/B/n test », is a type of experiment using two or more variants of the same web page.Â
Variant A is the original.
Variants B to n , each contain one or more elements that are modified from the original.
In this example, A represent the original title (in red) and B is the new title in test (in green).
With Lengow you can easily run « A / B » tests with your Google Shopping feed!
=> You wish to test the impact of a new title on a group of products?
This is now possible through Lengow using Custom Labels from your Google Shopping feed. Indeed these Custom Labels, common to the products (according to your key performance indicators) will allow you to structure and optimize your Shopping campaigns.
How can I set up my test campaign on Lengow ?
First, select the products you want to integrate into your A / B test by isolating them in « test groups » (products or product categories for example).
Once you know which products you want to test, you can set up a rule in Lengow to assign the Custom Label « test » to this product group and define your optimization actions.
- The custom label will help identifying your test products for the statistical analysis in Google Adwords
- The condition of your rule must retrieve all products that are tested as below.
Ex RULE A : I want to test my title by filling in the following attributes: Product name + brand + color.
In attribute Custom_Label_0 I add a new condition.
I name my rule « Test A Custom label 0 ». Then I retrieve my « test products » thanks to their category by setting up the following condition :
When my CATEGORY contains Sacs & Fourre tout > Etuis
Then I add the following actions : First one is for the attribute Custom_Label_0 that becomes « testA » :
Then Custom_label_0 becomes testA
The second action is for the Title that becomes « Product name + brand +color » :
Then Title becomes $NAME_PRODUCT - $MANUFACTURER - $COULEUR
Ex RULE B : My second test must keep the same attributes in the title but in a different order: the brand + the color + Product name.
This time I'm using another Custom_Label. It will help Google statistics to get the information separately from Custom_Label_0.
I choose the attribute Custom_Label_1.
I name my new rule: « Test B Custom label 1». Then I get the same « test products » thanks to their category by setting up the same condition as for the rule A:
When my CATEGORY contains Sacs & Fourre tout > Etuis
Then I add the follwoing actions : First one is for the attribute Custom_Label_0 that becomes « testB » :
Then Custom_label_0 becomes testA
The second action is for the Title that becomes « brand + color + product name » :
Then Title becomes $MANUFACTURER - $COULEUR - $NAME_PRODUCTÂ
Analysis
Once your A / B testing campaign is launched, you will be able to measure the performance of your campaign in the « Dimensions » tab of the Google AdWords platform.
With the example above you will see if the product titles defined with the attribute Custom_label_0 were more successful (audience, performance) than the titles of the attribute Custom_label_1 or even from the original titles.