Cordial provides a way to test message content with split tests. They are called experiments and are more powerful than traditional A/B tests because they allow testing of multiple variants with or without a champion. Split tests without a champion also allow for a control variant.
The method of determining the champion is a probabilistic evaluation of the conversion metric.
Cordial allows testing of 2 conversion metrics:
- Unique opens
- Unique clicks
Future enhancements will include other KPIs (orders, revenue, custom conversion events etc.).
Experiments may be placed in the subject line (via the UI) or message body (using smarty).
Note: multiple experiments may be run concurrently in batch messages, but they must use same testing strategy and populations (ex: strategy="split" or strategy="split-with-champion"). If you set your subject line experiment to "split with champion" your body experiment must also use the same strategy and populations.
Multiple experiment examples:
- Subject experiment AND body experiment (must use same strategy and populations for each).
- Multiple body experiments (must use same strategy and populations for each).
Note: Each experiment is evaluated uniquely and are not connected in any way. There is no a way to ensure a particular subject line variant gets delivered with a particular body experiment variant.
In the next article learn about subject line experiments in batch messages.