See how a home improvement and trade supply retailer used a geo experiment to measure what Connected TV really delivered, and how far it could scale before returns fell off.
A 42-day geo incrementality test showed that Connected TV returned $3.14 in incremental revenue per dollar for a home improvement and trade supply retailer, over twice YouTube’s return at matched spend. New customers drove most of the lift. Calibrating the brand’s marketing mix model with the result showed room to scale CTV from $2,400 to $12,500 a day, worth an estimated $12.1M a year. Here’s how the test worked.
Business Challenge
Connected TV was part of the brand’s media plan, but it sat outside its core measurement model. The team could see how many households its CTV campaigns reached and how often. Reach and frequency, though, don’t show whether those impressions changed what anyone bought. Without that, CTV’s budget was hard to defend and impossible to size. The team needed answers to two questions:
How much revenue does CTV actually add?
How far can spend grow before the next dollar stops paying for itself?
Objectives
Measure CTV’s incremental revenue with a live experiment, since the channel had no read in the existing model.
Compare CTV’s return with other video spend, starting with YouTube at a matched spend level.
Bring CTV into the model with a calibrated response curve, and find how far spend could scale profitably.
How LiftLab Measured CTV and Mapped Its Response Curve
The Incrementality Testing Suite ran a single geo experiment designed to answer both questions.
Go Dark with Pacing geo test: For 42 days, CTV ran in a matched treatment cell covering 14.87% of markets and went dark in a control cell covering 14.63%, at an average of $16,011 a day.
Lift by customer type: The test read incremental lift for new and existing customers separately, to show who CTV was actually moving.
Matched-spend benchmark: CTV’s iROAS was compared with YouTube’s at the same spend level, so the two video channels could be judged on equal terms.
Calibration through the Trust Engine: The results were fed into the brand’s Agile MMM, giving CTV its first experiment-calibrated response curve.
Results & Business Impact
CTV returned $3.14 in incremental revenue for every dollar spent at the test’s average of $16,011 a day.
That’s more than twice YouTube’s return at matched spend ($3.14 compared with $1.35).
New customers drove most of the lift, at 5.6% compared with 0.6% for existing customers.
The calibrated response curve then showed how far CTV could scale.
Raising daily spend from $2,400 to $12,500 is projected to add about $12.1M in incremental revenue a year.
Under a 33% margin assumption, the curve puts the upper limit at about $28,000 a day, CTV would be worth an estimated $25.5M in incremental revenue a year.
The brand now has a measured return for a channel its model couldn’t previously see, and a sized plan for how far to take it.
Key Takeaways
Reach shows exposure, not impact. Reach and frequency tell you who saw an ad. Only an experiment tells you whether it changed what they bought.
Bring unmeasured channels into the model. A channel the model can’t see never gets a fair share of the budget. One well-designed experiment can give it a calibrated response curve.
Compare channels at matched spend. Returns change as spend changes, so comparing two channels at different spend levels can make either one look better than it is.
Judge a channel on the job it does. A 5.6% lift among new customers against 0.6% among existing ones says CTV’s value here is acquisition, and it should be measured on new-customer outcomes.
Scale in steps beyond the tested range. The $12,500/day recommendation sits inside the spend range the test covered. The $28,000/day ceiling sits well beyond it, so approach it in stages and re-test along the way.
Methodology note: iROAS (incremental return on ad spend) is the incremental return observed during the test window, comparing markets where CTV went dark (control) with markets where it ran (treatment). mROAS (marginal return on ad spend) is the estimated return on the next dollar at a given spend level, read from the calibrated response curve. Results reflect one anonymized customer experiment run in Q3 2026. Individual results vary by category, baseline demand and media mix.
Our customers’ names are kept anonymous to protect their confidentiality.
Do you know what your CTV spend is really adding?
FAQs about the CTV Incrementality Case Study
What is a CTV incrementality test?
It measures how much revenue Connected TV advertising causes that wouldn’t have happened otherwise. In a geo test, CTV is paused in a set of control markets and kept running in matched treatment markets. The difference in outcomes between the two groups is the incremental lift.
What is the Go Dark with Pacing methodology?
Go Dark with Pacing is LiftLab’s experimental method for measuring incremental media impact. It pauses delivery in selected control markets while maintaining normal delivery elsewhere, allowing measurement of the difference in business outcomes.
Why compare channels at matched spend?
A channel’s return falls as its spend rises, because each extra dollar reaches less responsive audiences. Comparing two channels at the same spend level removes that effect, so the difference you see comes from the channels themselves.
How do you find the right spend level for a channel like CTV?
A response curve shows how incremental revenue changes as spend rises. Reading the return on the next dollar (mROAS) along that curve shows where added spend still pays back and where it stops, which tells you how far to scale.
How does LiftLab bring a channel like CTV into its MMM?
LiftLab’s Incrementality Testing Suite runs a geo experiment on the channel, and the <a href=”https://liftlab.com/platform/real-time-mmm-intelligence/”>Trust Engine</a> feeds the result into Agile MMM to calibrate the channel’s response curve. Every budget recommendation that comes out of the model goes to your team to act on.





