Incrementality Reporting

Incrementality reporting helps you understand the true impact of your advertising. Specifically, how many conversions were caused by your campaign, beyond what would have happened naturally.

Traditional attribution can overstate performance by crediting conversions that may have occurred anyway. Incrementality adds a critical layer of insight by isolating net-new outcomes driven by your ads.

Incrementality reporting answers key questions like:

At a high level, results are based on comparing two groups:

The difference in performance between these groups represents your incremental impact.

The Incrementality dashboard is designed to help you quickly interpret results and take action.

A core visualization compares:

This shows how audiences performed with and without ad exposure. The gap between these rates is the foundation of incrementality.

Lift represents the increase in conversion rate driven by advertising.

Example:

This metric translates lift into real business outcomes:

This is often the most actionable metric for evaluating campaign impact.

The Likelihood Score shows how much more likely users are more likely to convert after being exposed to your ads, compared to those who were not exposed.

It is caculated by comparing the conversion rates of the two groups:

This is presented as a multiplier (or index), making it easy to interpret relative impact.

​Example:

How to interpret:

In the example above, a 1.5x Likelihood Score means that users who saw your ads were 50% more likely to convert than those who did not.

Incrementality results are evaluated for statistical significance, helping you understand whether observed differences between exposed and control groups reflect real impact rather than random variation.

In the dashboard, results are clearly labeled as either:

These ensure that observed lift is not due to random variation and reflects true campaign impact.

How this is determined

Jamloop uses a two-proportion z-test, a standard statistical method for comparing conversion rates between two groups:

This approach evaluates whether the difference in conversion rates is large enough to confidently attribute it to the campaign. We apply a high-confidence threshold (99%) to reduce false positives, meaning results are only marked as significant where there is strong evidence of real impact.

Incrementality reporting is best used to guide strategic decisions, such as:

Incrementality answers “Did this work?”, while attribution helps answer “Where should I optimize?”

Incrementality is measured using a controlled experiment:

Because these groups are defined upfront and drawn from the same eligible audience, they represent a fair, like-for-like comparison. This ensures that both groups had a real opportunity to be exposed to ads.

This approach differs from methods that create control groups retrospectively (often called "synthetic controls"), where users who did not see an ad are grouped after the campaign ends. Those approaches can introduce bias, for example: by including households that were never actually eligible to receive an ad.

By establishing a true control group before the campaign begins, differences in outcomes can be reliably and causally attributed to advertising.

To ensure accurate and statistically meaningful results, experiments require sufficient scale, duration, and proper setup.

As a general guideline, campaigns should reach enough households to produce consistent conversion activity across both exposed and control groups.

Incrementality relies on comparing two large, statistically comparable groups. Adequate scale and clean setup ensure: