How retargeting works
A tracking signal, usually a pixel or a server-side event, records who visited a page, added to cart, or engaged with your content. Those people are grouped into audiences, and ads follow them across the platform. Because these users already know you, click-through rates are high and cost per click is low, which is why retargeting reliably posts the best-looking numbers in most accounts. The question that matters is not whether the numbers look good, but whether the ad caused the sale.
The over-credit problem
Retargeting reaches people who were already on their way to buy, so it takes credit for conversions that would have happened anyway. Last-click attribution makes this worse by handing the final touch a disproportionate share of the reward. The result is a campaign that looks like a hero on the dashboard while its true incremental value may be modest. This is the single most misread line in most ad accounts, and the reason a great retargeting ROAS should invite a test rather than a budget increase.
Using retargeting well
Retargeting works best as a closer with sensible limits, not as a volume engine. Segmenting by depth of intent, such as cart abandoners versus casual browsers, lets you spend where the incremental case is strongest. Frequency discipline matters, because a small audience fatigues fast and heavy repetition annoys the very people most likely to buy. The honest way to size it is to hold a slice back and measure the lift, then fund it to the point where the incremental return, not the reported one, still holds.
How iClick uses retargeting
iClick runs retargeting as a measured closer rather than a headline performer. A strong reported ROAS on a retargeting audience is treated as a prompt to pressure-test with a holdout, because the warm audience inflates the claim. Budget is set against the incremental result the test reveals, and frequency is capped by fresh creative rather than brute repetition, so the closer keeps working without burning the audience it depends on.
