Ghost ads methodology
Ghost ads are an experimental design for measuring ad effectiveness. Rather than showing a placebo ad to the control group, the system records which control users would have been served the test ad had they been in the treatment group, then compares only comparable users. This isolates the ad's incremental effect while avoiding the cost and bias of serving placebo creative.
What this means
A classic challenge in ad experiments is that the control group cannot see the test ad, but you still need to know which control users would have been served it — otherwise you compare exposed users against a control that includes people the ad never could have reached.
Ghost ads solve this by running the ad-serving logic for the control group and logging where the test ad would have appeared, without actually showing it (or showing nothing in that slot). The comparison is then restricted to the comparable subset: treatment users who saw the ad versus control users who would have. This removes a major source of selection bias.
Why it is preferred over placebos
An older approach uses public-service announcements (PSAs) as placebo ads for the control group. Ghost ads improve on this by not spending budget on placebo creative and by avoiding any effect the placebo itself might have. The methodology was introduced in academic and industry research and underpins lift products in several platforms.
Like all incrementality methods, ghost ads measure aggregate causal effect over a window, not per-touch credit, and require sufficient scale. They are an experimental technique, not a reporting attribution model.
- Logs would-have-been-served control users for fair comparison
- Avoids paying to serve placebo (PSA) creative
- Reduces selection bias versus naive exposed-vs-all comparisons
How it appears in analytics and logs
A ghost-ads comparison limits both groups to users the auction would have reached, so the measured difference reflects the ad's effect rather than differences in who was reachable.
Diagnostic use case
Use ghost ads when you want incrementality measurement without paying to serve placebo ads, by tagging the would-have-been-exposed control users for an apples-to-apples comparison.
What WebmasterID can help detect
WebmasterID does not run ad auctions, but its first-party conversion events can be the outcome measured when a platform reports ghost-ads-based lift.
Common mistakes
- Comparing exposed users to a control that includes unreachable users.
- Treating ghost-ads lift as per-channel attribution credit.
- Assuming any control design removes all bias without enough scale.
Privacy and accuracy notes
Ghost ads operate on aggregate exposure-eligibility flags, not personal profiling for re-identification. This page is educational, not legal or statistical advice.
Related pages
- PSA control group testing
A PSA (public-service announcement) control test is an incrementality design where the control group is served unrelated placebo ads instead of the test campaign. Because both groups receive an ad impression, exposure conditions are similar, and the difference in conversions estimates the test campaign's incremental effect. It is an older alternative to ghost ads.
- Conversion lift studies
A conversion lift study randomizes users into a group eligible to see ads and a control group held out from them, then compares conversion rates between the two. The difference estimates incremental conversions — those caused by the ads rather than ones that would have occurred anyway. Major ad platforms offer lift studies as a counterfactual check on attributed conversion counts.
- Incrementality testing: what attribution cannot tell you
Incrementality testing measures the lift a channel actually causes by withholding it from a control group and comparing outcomes. It answers the question every attribution model dodges: would this conversion have happened anyway? It is causal where attribution is merely correlational, but it requires deliberate experiment design.
- Attribution analytics
Use first-party outcomes alongside platform lift methods.
Sources and verification notes
- Johnson, Lewis & Nubbemeyer — Ghost Ads (Journal of Marketing Research)Introduces the ghost-ads counterfactual design for ad measurement.
Last reviewed 2026-06-24. Facts are checked against primary/official sources where available; uncertain specifics are marked “Data not yet verified” rather than guessed.