Our online store's retargeting campaign reports an 8x return on ad spend, while prospecting is closer to 2x. The obvious recommendation is to move more budget into retargeting, but I am uneasy about that comparison.
The retargeting audience includes people who recently added products to their carts. Some probably would have purchased through email reminders or returned on their own. The platform also counts purchases under its view-through attribution window.
What would make this a fair test?
I want to measure incremental sales, not just purchases that happened after an ad impression. Would a randomized retargeting holdout test be appropriate for a store with only a few hundred orders per month?
How should we handle customers who also receive email campaigns, and should the decision metric be additional revenue or additional contribution margin after advertising costs? I would also like to understand how to interpret a test that does not produce a clear result.
