Incrementality Testing

How It Works

1. Experiment Design

To set up a new experiment, the AI expert will guide you step-by-step through the following key decisions:

  • Experiment Goals: Clearly define the objective of your experiment
  • KPI Selection: Choose the right metrics to measure success
  • Power Analysis: Determine required budget, test duration, and optimal test & control region selection to achieve the desired statistical power

The design splits the US into test and control regions that historically move together, so the comparison is fair from day one. Tests can run at the state level, but fine-grained commute zones are generally recommended — better-matched regions, smaller holdouts, and more sensitive measurement.

US map of commute zones colored as test regions (red) and control regions (blue)
An experiment design: power analysis selects well-matched test and control regions across the US.

2. Launch Experiment

Once the experiment is designed, launch it directly from Maxma with one click: connect your ad accounts and Maxma applies the geo targeting changes to your campaigns automatically — and reverts them when the test ends. See Ad Platform Integrations for supported platforms. For other channels, marketers can implement the test manually using the pre-defined experiment specification from the design step.

3. Results Analysis

When the experiment concludes, marketers return to the Maxma platform to analyze the results and gain insights. See Deliverables for what the results include.

Previous
Deliverables