Why retargeting looks so strong

People who already showed interest or purchase behavior have a high baseline probability of buying again. If they see an ad and then convert, an attribution system can make the tactic look extremely efficient.

Standard ROAS lacks the counterfactual

The effectiveness question is what would have happened without the advertising. Without a control design, attribution can assign credit for organic or otherwise expected sales.

High CPMs can buy existing demand

Precise targeting is technically possible but not automatically economically optimal. If media mainly reaches people who would have purchased anyway, reported efficiency can be high while incremental impact is small.

Growth requires new or additional demand

Fou argues for greater emphasis on people who are not already highly likely to buy, including audiences who are not yet aware of the brand or product.

Read attribution and incrementality together

Attribution describes which touchpoint receives credit under a model. Incrementality estimates the additional causal effect created by the intervention. Effective media governance should not confuse the two.

Methodological note

This article describes principles and questions. Specific fraud, bot or quality rates cannot be transferred directly to individual campaigns without understanding methodology, scope and data basis.

SOURCE / ORIGINOriginal article by Dr. Augustine Fou: Targeting and Retargeting – Looks Like It’s Working, but Worst for Incremental Outcomes Original publication: 2025-04-07