What it means
Incrementality tests answer the question attribution cannot: how many of the conversions credited to a channel would have happened anyway? Common test designs include geo-experiments (markets are split into test and control), platform-native lift studies (e.g., Meta Conversion Lift, Google Lift), and ghost-bid or pre-test holdouts. Incrementality is now considered the gold standard for measuring channel value because it isolates causal effect rather than correlation, and it is unaffected by attribution windows, last-click bias, or cookie deprecation.
How Coegi approaches it
Incrementality testing is built into every Coegi engagement above a defined spend threshold. Tests are designed before campaigns launch — not bolted on afterward — so the resulting data is statistically valid and actionable for budget allocation decisions.
