Experiment / All other cases for placements

Hi ​@quentin,

Thanks for reaching out. I’m happy to help.

When a customer is enrolled in an Experiment, the experiment’s Placement Offering map is what gets used, and it won’t “fall back” to your normal Targeting rule for any Placement that isn’t explicitly defined in the experiment. That’s why, if you want to only test/override the promo placement, you’ll still want to define the Offering for every Placement your app uses in the experiment (Control + Treatment), keeping them identical except for promo in the Treatment.

To reduce the errors/make it less cumbersome, consider duplicating an existing experiment configuration and tweaking just the one placement you’re testing.

Hope this helps, let me know if you have any other questions.

Best,

Hussain

That's a major regression from what we had before, is there any plan to bring back the override of only 1 placement? 

Hi ​@quentin,

Thanks for following up. You can use the "All other cases" placement in experiments, the offering selected there will be returned to any placement not defined in the offerings table for the experiment. This can make it a tad bit easier to configure experiments.

However, as I’ve mentioned before when a customer is enrolled in an Experiment, the experiment’s Placement Offering map is what gets used, and it won’t “fall back” to your normal Targeting rule/Default offering for any Placement that isn’t explicitly defined in the experiment.

I’ve shared your feedback with the product team, but currently do not have a definitive timeline on when/if the “override just one placement” option will be introduced for experiments.

Hope this helps, let me know if you need help with anything else.

Best,

Hussain