Marketers Must Test Incrementality Before Peak Spending

Marketers need to test incrementality before peak spending starts to avoid over-crediting certain channels, MarTech reported. The goal is to determine which channels generate incremental revenue. By doing so, marketers can make informed budget decisions.

During peak season, a significant portion of annual revenue is generated in a short period. For instance, Cyber Week 2025 saw a record-breaking $44.2 billion in online spending. As a result, over-crediting a channel can be costly.

Holdout tests are controlled experiments that measure the incremental revenue generated by a particular channel. This is done by switching off the channel for a portion of the audience and comparing the results to the rest of the audience.

To design a holdout test, marketers should choose a channel and switch it off for part of their audience. The test should be run for a sufficient amount of time to generate useful data. The results will show whether the channel is driving incremental revenue or not.

When running a holdout test, it is essential to keep the control group clean. This means switching off links, blocking coupon extensions, and informing key publishers about the test. Any potential contaminants, such as competitors' promotions, should be logged and considered.

The test results should be compared to previous periods and years to determine the incremental revenue generated by the channel. This will help marketers make informed decisions about their budget allocation. By moving from return on ad spend to incremental return on ad spend, marketers can get a clearer picture of which channels are driving real revenue growth.

Marketers should run holdout tests before peak season to get actionable data. This means starting the test early enough to get results before peak spending begins. It is also crucial to audit all links, tags, and pixels before running the test to ensure accurate data.