Interactive publication
Back to the catalogueAdvertising semanticsCase 21 of 22
Treatment/control contamination
In a geo-lift experiment, cells assigned to control can still receive treatment exposure via platform audience expansion, ad-adjacent delivery, or a resident's movement into a treatment cell, biasing the measured lift toward null.
Class
Case to account for
Affects
2 conversions
01
Detection
Non-zero measured impression/exposure delivery within cells designated as control.
02
Mitigation
01
Use a geographic buffer/no-man's-land between treatment and control cells to absorb spillover
02
Independently verify zero-delivery in control via platform reporting or third-party exposure logs, not assumption
03
Affected conversions
Cell systemcell system
OPENAdmin polygonsH3 (weighted crosswalk)
OPENMore in Advertising semantics
Bidstream truncationCarrier NAT / multi-family IP poolingCorridor boundary oscillationCross-device location stitchingDeclared resolution exceeds input accuracyDestination interestExcluded-area leakageFrequency across overlapping targetsGPS noise / boundary oscillationGPS-derived location accuracyHome location inferenceIP-derived locationLocation confidence scoringLocation interest versus presenceLookback window sensitivityMixed-provenance channel bundlePhysical presencePooled location sources in delivery logsPublisher-declared locationRecent presenceWork location inference