GLOSSARY

The incremental increase in conversions caused by an intervention, estimated by comparing outcomes for an exposed treatment group with a comparable unexposed control group.

Conversion lift helps decision-makers distinguish conversions caused by advertising from conversions that would have happened anyway. It supports budget allocation and incrementality analysis beyond attribution reports.

An advertiser runs a user-based holdout study in which an eligible control group is withheld from campaign exposure. The team compares treatment and control conversions, reviews confidence and minimum-detectable-effect assumptions, and uses incremental cost per acquisition to inform the next budget decision.

Low conversion volume, contamination between groups, short study windows, seasonality, non-compliance, and poor randomisation can make results inconclusive or biased. Lift applies to the tested population, intervention, and period; it should not be generalised automatically or confused with attributed conversions.

A lift study assigns eligible users or geographic areas to treatment and control conditions, observes conversions over a defined period, and estimates the difference attributable to exposure. Absolute incremental conversions, relative lift, uncertainty, and incremental cost per action should be interpreted together. Randomisation or a defensible quasi-experimental design is required for causal interpretation.