What is incrementality

​

What is incrementality compared to attribution?

Attribution models and incrementality testing both aim to measure marketing impact, but they do so in fundamentally different ways.

​

Attribution: Assigning credit to touchpoints

Attribution models (such as last-click, first-click, or data-driven attribution) attempt to allocate credit for conversions to different marketing touchpoints. These models help marketers understand which channels or campaigns are involved in the customer journey.

Example: A customer clicks on a Google ad before purchasing a product. Attribution models may assign credit to the ad, but the customer might have purchased the product anyway—even if they had never seen the ad.

​

Limitations of attribution models

• They assume correlation equals causation. Just because a user clicks on an ad doesn’t mean the ad caused the purchase.
• They don’t account for organic behavior. Some customers would have converted without any marketing exposure.
• They struggle with cross-device tracking and walled gardens. As users move between devices and platforms, attribution models often lose visibility.

​

Where does MMM fit in?

Marketing mix modeling (MMM) is another approach used to measure marketing impact, particularly at a broader level. It uses statistical models to analyze historical data and estimate how different marketing channels contribute to overall sales.

​

Strengths of MMM

• Provides a holistic view across all channels, including online and offline marketing.
• Doesn’t rely on user-level tracking, making it privacy-friendly.
• Useful for long-term budgeting and strategic decisions.

​

Limitations of MMM

• Requires large amounts of historical data and works best with longer timeframes.
• Less precise for tactical optimizations, since results are often aggregated and not at the individual campaign level.
• Difficult to measure rapid changes, as updates require recalibrating the model.

​

How incrementality testing enhances MMM

While MMM provides a big-picture view, incrementality testing offers a high-precision measurement of specific campaigns or channels. Running both in parallel allows marketers to:

• Validate and calibrate MMM results with real-world experiments.
• Use incrementality testing for tactical decisions (e.g., testing a new channel or campaign change) and MMM for strategic planning.
• Ensure that media mix decisions are based on causal impact rather than just historical correlations.

​

Summary

• Attribution models track and assign credit but don’t prove causality.
• MMM estimates overall channel impact but works best at a high level.
• Incrementality testing directly measures the true lift from marketing spend through controlled experiments.