Retargeting, also called remarketing, reaches users who have already shown intent. Because the audience is pre-qualified by its own behaviour, conversion rates sit well above prospecting and cost per acquisition well below. It is also the easiest place in all of media buying to fool yourself, for exactly the same reason.

How it works

A user action is recorded through a pixel, an SDK event, or an uploaded customer list, and that user joins an audience segment. Campaigns then target the segment, often with creative tied to the specific product the user looked at. Segments carry a membership window so that intent expires, since someone who browsed six months ago is no longer a warm prospect.

Why it matters

Most people do not buy on a first visit. Retargeting exists to recover the majority who left, and on reported numbers it is usually the best-performing line in the account. The catch is that it is addressing people who were already close to converting, so a meaningful share of what it claims would have happened regardless.

The attribution trap

Retargeting takes credit for conversions that were already in motion. A user who added to cart, got distracted, and returned the next day would often have returned anyway, but if a retargeting ad appeared in between, the platform books the sale. This is why mature advertisers run incrementality testing on retargeting before scaling it: attributed ROAS frequently looks excellent while incremental ROAS looks ordinary.

Retargeting and rewarded placements

Rewarded environments approach the same problem from the opposite direction. Instead of following a user around until they come back, an offerwall presents the offer to someone who has opened it deliberately to find something worth doing. The intent is declared rather than inferred, which sidesteps both the frequency problem and much of the attribution ambiguity.

Retargeting vs. prospecting

Prospecting reaches people who have never heard of the brand and is judged on whether it creates new demand. Retargeting reaches people who already have and is judged on whether it converts existing demand. The two cannot be compared on the same efficiency metric, because retargeting will always look better and is always working with an easier audience. Accounts that shift budget from prospecting to retargeting on reported ROAS tend to see efficiency climb and total revenue flatten, which is the signature of a funnel with nothing new entering it.

Common mistakes to avoid

• Running it without a frequency cap. A small audience and an open budget produce punishing repetition, and frequency capping is the fix.

• Forgetting to exclude converters. Advertising a product to someone who already bought it wastes budget and reads as incompetence.

• Leaving membership windows too long. Stale intent converts like cold traffic but costs like warm.

• Scaling on attributed ROAS alone. Without an incrementality read, the budget may be buying conversions it did not cause.