When using Agillic, it's likely that you'll plan initiatives such as more dynamic content, improved layout, and styling, or special offers and discounts for your recipients.
As part of implementing such initiatives, you may want to test the effectiveness of your changes. You can do this by comparing recipients who received these improved outbound communications with a so-called control group who didn't receive the same outbound communication.
When testing on control groups in Agillic you should follow:
- Agillic's Best Practice for Test on Control Groups
Agillic's Best Practice for Test on Control Groups
Control groups shouldn't be confused with terms like Target Groups or Static Target Groups.
Our recommendations for control groups depend on what you want to happen to the control group during the test period.
No Outbound Communication to Control Group
If you have a new type of campaign, you may choose not to include certain recipients - your control group - in the Target Group for your campaign. This way, the control group wouldn't receive the outbound communications for this campaign during the test period.
To remove a control group from your Target Group, simply add a Person Data condition based on the Fixed Person Data field 'RANDOM'. We recommend you create both a Target Group of the total number of recipients for the campaign and another Target Group that contains the control group to be removed.
Once you've executed the campaign, you can compare the result for the recipients who were part of the campaign, to those in your control group. You'll likely report on this outside Agillic as this comparison will likely measure an external factor like the number of purchases in a webshop or a satisfaction score.
Other Outbound Communication to Control Group
For some campaigns, you may still need to communicate to your control group, even if the content may be more static or not contain the same initiative the main Target Group sees. If this is the case, we recommend you make a Random Split Flow for sending your different outbound communications.
Exporting the Target Group and Control Group
In order to know who was part of the campaign and in the control group, you should create an Export Flow for the total campaign Target Group and include the RANDOM Person Data. You should then use the Fixed Person Data Field 'RANDOM' to make the same condition in your external reporting tool, in order to split recipients into the control group and the participants of the main campaign.
Suggested Next Steps
- All You Need to Know About Data Models
- All You Need to Know About Target Groups
- All you need to know about Conditions
- How to Create Conditions