MEASURE THE TRUE LIFT OF YOUR MEDIA WITH REAL-WORLD EXPERIMENTS
Measure media lift with the precision only the most trusted scientific method can deliver—randomized control trials—and calibrate your attribution model for maximum accuracy.


%
INCREASE IN ADVERTISING ROI
%
HIGHER ROI THAN MMM ALONE
Join over 200 marketing teams making data-driven decisions with confidence

What makes incrementality experiments so powerful is their ability to uncover true causal relationships—rather than misleading correlations. For example, if you announce a promotion to your email list, some subscribers will search for your brand and click on your Google brand ads. This creates a strong correlation between brand ad spend and sales—but the sales weren’t caused by the ad spend. A third factor—a confounder—drove both: the promotion.
Incrementality testing controls for these confounders. If the promotion goes out to all regions, but only some regions see the ads, you can isolate the effect of the ads themselves—separating their impact from the promotion’s. That’s how you get to the *true* lift.

How It Works

STEP 1: SELECT THE EXPERIMENT TYPE
Choose between a holdout test (to see how much sales drop when you pause a channel) or a scale-up test (to see how much sales rise when you increase spend). While holdout tests measure ROAS at your current budget, scale-up tests measure ROAS at a higher spend—and calculates the incremental ROAS between the two.
STEP 2: DESIGN YOUR EXPERIMENT
We’ll automatically select the best regions and experiment duration to ensure a large enough sample size for statistically significant results. Regions are chosen to accurately represent your market while accounting for only a small share of revenue—to prevent interfering with your most profitable areas. Because each test runs in a separate geographic market, our platform supports up to three simultaneous tests.
STEP 3: ANALYZE THE RESULTS
Our platform shows how much revenue changed during the experiment—and whether that change is statistically significant. We use the synthetic control method (SCM) to estimate the causal impact of your campaign, comparing actual revenue to the counterfactual: what would have happened without the intervention, based on a weighted average of the control regions.
STEP 4: CALIBRATE YOUR MEDIA MIX MODEL
One of the biggest advantages of incrementality testing is the ability to calibrate causal inference tools like media mix models. While attribution models can work without incrementality tests, companies that combine both see a 59% greater increase in ROAS (35% vs. 22%). When you run incrementality tests with Data Speaks, your attribution models are automatically calibrated for maximum accuracy.
WHAT OUR CLIENTS ARE SAYING
Ready to see Data Speaks in action?
Book a demo to explore the platform and decide if it’s right for you.
Not ready for a demo? Contact us at hello@dataspeaks.ai
All Your Data in One Place
You rely on data to make decisions, but who has the time to log in daily to a dozen different platforms?
Google Analytics, Shopify, Amazon, advertising platforms, email, SMS, and more.
You need all your data in one place to make informed decisions in real-time and lead your brand to success.