Use case: Testing two WhatsApp campaign variants
Compare two versions of a WhatsApp campaign to see which performs better.
For example, you might want to test:
Two different message variations
Two promotional offers
Two calls to action
Two template designs
Use a Program with a Random split node to divide contacts between campaign variants. After sending the campaigns, compare the results to identify which variant performs best.
Solution overview
Use this approach to test WhatsApp campaign content with a randomly distributed audience.
Use a Program to:
Split contacts into separate groups using a Random split node.
Send a different WhatsApp campaign variant to each group.
Review campaign performance and identify the better-performing variant.
Example: Test two WhatsApp campaign variants
You're promoting a seasonal offer and want to test which message generates more engagement.
Create two separate WhatsApp campaigns:
Variant A
“Save 20% this weekend. Shop now.”
Variant B
“Weekend flash sale. Get 20% off before midnight.”
In this example, each WhatsApp campaign represents a test variant.
You then use a Program with a Random split node to divide your audience between the two campaign variants.
50% of contacts receive Variant A.
50% of contacts receive Variant B.
After both variants have been sent, compare the results manually to determine which version performed best.
Step-by-step guides
1. Create the workflow
Go to Automation > Programs.
Build a simple program that manages WhatsApp campaigns.
Drag a Random split node onto the canvas and connect it to the start node.
Configure the audience split percentage. For example, 50% and 50%, or use a smaller test segment such as 10% and 90% if you plan to send the better-performing variant to the remaining audience later.
Add a WhatsApp campaign node to each branch.
Select a different WhatsApp campaign for each branch.
Select Save.
2. Test the program and activate
Select Validation to check if the program is correct.
Select Activate to start the program.
Marketing sends the campaigns according to the program's Enrollment scheduling settings.
The program randomly distributes contacts between the branches. Each audience segment receives a different WhatsApp campaign variant.
3. Analyse results
After the campaigns have been sent, review the reporting for each campaign in Analytics > Reports > WhatsApp to identify the better-performing variant.
Depending on your goal, compare metrics such as:
Delivery rate
Read rate
Link clicks
Use the results to optimise future WhatsApp campaigns.
4. Apply your results
Use the results from your test to inform future WhatsApp campaigns.
The example in this article uses a 50/50 audience split to compare two campaign variants. After identifying the better-performing variant, consider using the insights from the test in future campaigns or when creating new WhatsApp message templates.
You can also repeat this process to test additional elements, such as:
Different promotional offers
Alternative calls to action
Different message wording
Template designs
Continuous testing helps you understand which types of WhatsApp campaigns perform best for your audience.
Optimise your testing strategy
Create a test audience that represents a small percentage of your total audience, such as 10%.
Keep the remaining contacts as a holdout audience.
Enrol the test audience into the Program.
Use a Random split node to divide the test audience between two WhatsApp campaign variants.
Compare the results to identify the better-performing variant.
Send the better-performing variant to the holdout audience as a separate WhatsApp campaign.
This approach allows you to test campaign content on a smaller audience before sending it to the majority of your contacts.