Lift testing is a foundational methodology for marketers seeking to prove the true impact of their media investments. This article provides a comprehensive overview of how lift testing works, how WorkMagic supports it, and how you can successfully run and interpret lift tests using our platform.

A lift test is a controlled experiment that measures the incremental impact of your advertising. It works by comparing a Test group (exposed to ads) to a Control group (not exposed) and quantifying the difference in outcomes (e.g., orders, sales, new customers).

Lift tests answer the question:

“What would have happened if we hadn’t run this channel/campaign?”

Unlike attribution models, lift tests isolate the causal effect of your advertising, helping you make smarter investment decisions based on actual performance rather than correlations.

Lift testing is a powerful framework for answering a range of strategic marketing questions. Below are common use cases that WorkMagic can help you address:

WorkMagic is a managed experimentation platform that handles:

WorkMagic supports automatic holdout setup and teardown via API for:

For non-integrated platforms (e.g. AppLovin, MNTN, etc), we provide step-by-step instructions and DMA files to guide your team through manual setup.

We ingest order and sales data at geo-level via:

Lift testing helps you evaluate the true effectiveness of your marketing efforts by measuring incrementality—the impact your ads had compared to a scenario where they were not shown. WorkMagic provides a step-by-step interface to configure and schedule a lift test.

1. Choose the Test Type

You can run two types of lift tests on WorkMagic:

2. Select the Primary Metric

Choose the key outcome you want to measure:

3. Select Test Country (Required) and Location Exclusion (Optional)

4. Select Test Channels

Available platforms and their holdout capabilities include:

You can run 2-cell or 3-cell test. In each test cell/group, you can holdout multiple ad platforms (no limitations on the number).

Decide the level at which you want the test to run:

WorkMagic automatically recommends geo-pairs for your lift test. You’ll see two test methodologies:

Both methods measure incremental lift with the same confidence; the difference is how they split your audience and the budget required.

For each geo pair, WorkMagic displays:

Click into a geo pair to see region details, alternative budget options, or set a custom test period (14–35 days).

If your selected channel doesn’t have enough historical data, the recommended daily budget will show as Not available.

In this case, click Estimation Hypothesis and manually enter a Cost per Order value. WorkMagic will use that input to generate budget and test period estimates.

This ensures you can still run a lift test, even for new channels or ad types without past performance data.

You can apply the holdout in two ways:

Choose your start date and time, the testing and cooling period will be automatically selected.

Before launching, confirm all details:

Once reviewed, click Schedule now to launch your lift test.

If you need further assistance at any step, please contact your WorkMagic data scientist or use the “Contact WorkMagic data scientist” link available in the interface.