How it works
Use a suitable experiment or justified causal model; ordinary attribution asks a different question.
Example
Example: A control group helps assess whether advertising adds purchases rather than merely recording them.
Common mistakes
A before-and-after change alone does not prove causation.
How can you check this in practice?
With suitable scope, compare test and control groups under predefined conditions. Check group comparability and whether other material changes affect the period. Conclusions need to include uncertainty and design limitations.
Distinctions and limitations
Attribution credits observed interactions; incrementality asks what advertising added compared with its absence. Recorded ROAS alone does not establish causally additional sales.
Official sources
Editorial responsibility: Biser Bozhinov · PPC Specialist
Updated: 05/11/2026
For a correction or question: office@z-consult.bg