Foundation Metric

Incrementality Testing

Measure true causal impact of your marketing. Understand what drives real growth versus what merely correlates with it.

The Attribution vs. Causation Gap

Attribution shows correlation: who touched the customer last

Incrementality reveals causation: which ads actually changed behavior

Standard analytics credit channels that are present, not causal

Most 'high-performing' channels are harvesting existing demand, not creating it

How Incrementality Testing Works

Core Concept

Compare two groups: one exposed to marketing (test group) vs. one not exposed (control group). The difference in outcomes reveals the true incremental impact.

Incremental Lift = Revenuetest - Revenuecontrol
Geo holdouts: Randomly select markets to receive no ads, compare revenue to active markets
Time-based tests: Pause spending in specific channels for periods, measure impact on conversions
Synthetic control: Use statistical modeling to create virtual control groups from historical data

What Incrementality Testing Reveals

Demand Creation Channels

  • • Prospecting ads
  • • Brand awareness campaigns
  • • Top-of-funnel content
  • • New audience expansion

High incrementality, often under-credited by attribution

Demand Harvesting Channels

  • • Retargeting ads
  • • Branded search
  • • Lower-funnel conversions
  • • Existing customer re-engagement

Low incrementality, often over-credited by attribution

Real-World Example

Facebook Retargeting Campaign

Last-Click ROAS
8.2x
MTA ROAS
6.5x
Incremental ROAS
1.9x
Insight: Geo holdout test revealed that 77% of attributed conversions would have happened anyway. The campaign was harvesting demand created elsewhere, not generating new revenue.

Stop Guessing. Start Measuring Causation.

SpendSignal runs continuous incrementality tests to show you what's really working.