Brand lift for Connected TV, without cookies or device IDs
CTV is one of the fastest-growing parts of the media plan and one of the hardest to measure. There are no cookies on the big screen, no clicks and patchy device IDs. Most campaigns finish with a reach figure and a completion rate. Feather adds the number brands actually need, whether the campaign move recall, attribution and consideration.
Request DemoWhy CTV has been hard to measure
Brand lift needs two groups, people who saw the campaign and people who didn't. On a TV screen there is no reliable way to tag either group, and running a survey on the same TV inventory would only ever reach people who were exposed. So most CTV campaigns have gone unmeasured on brand outcomes.
How it works
1.The campaign runs on CTV.
On streaming apps and Smart-TV channels.
2.The survey runs on mobile, in parallel.
That lets us reach both CTV viewers and people who don't watch those channels.
3.We build exposed and control groups from viewing behaviour.
Respondents tell us whether they stream on their TV and which apps and channels they use. People who watch the campaign's channels form the exposed group. Everyone else forms the control.
4.We read the lift per KPI.
Each result is the exposed response rate minus the control rate, reported as Significant or Directional.
Why we ask about viewing, not about the ad
Asking "do you watch this streaming app on your TV" gives a much more stable answer than asking "did you see this specific ad". We build the groups on the durable signal, then measure the brand questions against them. Because the groups come from what people tell us, we state the confidence level for every result and show how the result holds up under stricter and broader definitions. Where addressable signals exist, we move to a deterministic design.
What you get
Brand Lift Score
one number summarising the campaign's impact, where 50 is the European digital benchmark (indicative).
Lift per KPI
Ad Recall, Brand Recall, Attribution, Consideration and Favorability.
Confidence per KPI
Significant or Directional, never buried.
eCPLU
the effective cost per lifted user, when spend is shared.
Built for privacy
No cookies, no device IDs and no cross-site tracking. The method measures outcomes, not people, and is designed around GDPR from the start. It also keeps working as identifiers disappear, because nothing depends on them.
Where it runs
CTV Lift is running in Germany and Austria through our partnership with YOC, across streaming and Smart-TV channel apps, with more markets in the pipeline.
Who it's for
Brands and agencies buying CTV
who need a brand outcome they can defend, not just a completion rate.
Publishers and CTV sellers
who want to offer measured brand impact with their inventory. Measured outcomes are a reason to buy, not an extra cost.
FAQ
How do you know who saw the ad?
We don't track individuals. We ask which streaming apps and channels people use on their TV, and compare people who watch the campaign's channels with people who don't.
Is self-reported viewing reliable?
It's far more reliable than asking people to remember a specific ad. We also report the confidence level of every result and show alternative cuts, so nothing is overstated.
How much does it cost?
Studies are designed to sit under 5% of media spend. Talk to us for a quote for your campaign.
Does it work alongside mobile and display measurement?
Yes. CTV Lift uses the same KPIs and the same reporting as our other brand lift studies, so you can compare across channels.
Find out what your next campaign changed.
Request Demo