Brand lift methodology
Last updated
Last updated
Happydemics enables you to measure the performance of your campaigns across most media channels and industries. We achieve this by collecting respondents' opinions through short surveys, showing them the ad, and evaluating high- and mid-funnel KPIs to benchmark your performance against our extensive database.
At Happydemics, our methodology is built around Ad recall. Our approach focuses on comparing “Ad recall respondents” to “Non-Ad recall respondents” (the control group), distinguishing us from the more common “Exposed vs. Non-Exposed” approach used by many brand lift solutions.
By default, a brand lift study is conducted with 500 respondents: 250 Ad recall respondents and 250 Non-Ad recall respondents (the control group). If needed, you can increase the sample size to double or triple the number of respondents for an additional cost.
Ad exposure: a flawed metric
Even with advanced cookie-based retargeting methods, exposure does not guarantee that the ad was actually seen:
Users might visit a page but never scroll far enough to see the ad.
Apps running in the background can serve ads without the user being aware of them.
Ad blockers can prevent ads from appearing at all.
Ads shown on shared screens—like a TV in a household—don’t identify who in the room actually saw them.
These examples highlight why exposure doesn’t always equal visibility. With Ad recall, we bypass these limitations by focusing on whether the respondent not only saw your campaign but also remembers it through a specific channel.
Preparing for a post-cookie world
As third-party cookies are gradually phased out, exposure-based methods will become increasingly difficult to sustain. Our recall-driven methodology is future-proof, ensuring accurate insights even as tracking technologies evolve—something many exposure-based solutions won’t be able to claim for much longer.
For large-scale digital campaigns, creating a "clean" non-exposed control group is becoming increasingly difficult. Tracking inaccuracies can muddy the waters, leading to skewed comparisons between exposed and non-exposed groups. Our approach eliminates this issue by distinguishing groups based on whether or not respondents remember your ad or not, making analysis easier and more accurate.
Closer to reality, Happydemics uses the “probable exposure” concept
Happydemics takes a more realistic approach by using the concept of "probable exposure" to closely match your advertising goals and provide an accurate recall rate. Whether it’s retargeting pixels, ID batches, or geolocation data (latitude/longitude), you can provide us with these details to refine the definition of probable exposure. However, we can also gather recall respondents without relying on this type of data.
More Accurate Insights
Exposure-based brand lifts often include people who technically “saw” the ad but didn’t notice or remember it, leading to diluted results. By focusing on Ad recall, we provide a sharper understanding of how your campaign performed and where it can improve. For example:
Imagine a brand lift study showing only a small increase in consideration. Is it because your campaign wasn’t memorable? Or because it failed to persuade its audience? Traditional methods don’t provide clarity.
With Happydemics, you’ll know if the problem is recall rates or the creative’s ability to deliver the intended impact.
Creative Feedback You Can Use for Optimization
Respondents who recall your ad can also provide valuable feedback on the creative itself. They help us provide creative diagnostics based on KPIs such as Ad likeability, Interest, Clarity or Ad perception. This level of insight is not possible with standard exposure-based approaches. These diagnostics offer actionable recommendations to refine your creative strategy and improve the impact on future campaigns.
Reliable Benchmarks for Context
Our benchmarks span 12 channel types, 30+ formats, and countless industries and regions, helping you evaluate your results in a meaningful context. Unlike other brand lift studies, where uplifts aren’t always comparable due to difference in methodology, our benchmarks are built on a consistent foundation. This ensures that you always know where your campaign stands, not just in isolation but against relevant industry standards.
An independent and innovative solution
Solutions from major platforms such as Meta and Google are often seen as the industry gold standard. However, these platforms are not neutral—they rate campaigns they also run, making them both judge and jury.
At Happydemics, we act as an independent, unbiased third party, delivering insights you can trust. And while other platforms rely on their proprietary data, we provide a holistic view across all channels, helping you assess your campaign’s full impact—without restrictions.
Our recall-based methodology works seamlessly across both online and offline channels, enabling easy comparison of results from multi-channel campaigns. This approach gives you a clear understanding of how each channel contributes to your performance, helping you optimize your media mix.
In contrast, the “Exposed vs. Non-Exposed” approach often varies by media type and, in some cases, cannot be applied at all. A brand lift study typically focuses on one channel. For instance, if your campaign runs across three different channels that you want to compare, you would need to conduct three separate brand lift studies.
Each brand lift includes a contextualized Ad recall question, designed to maximize the probability of capturing exposure on a given channel.
For example, for a DOOH campaign:
In addition to the contextualized Ad recall question, you can enhance the probability of exposure based on the media used.
The Happydemics Pixel: Potential exposure is tracked using the device’s IP. Exposed individuals can then be surveyed on other devices connected to the same wifi: desktop, mobile, tablet.
​POI (Points of Interest): Respondents are qualified by their presence within a given radius of a network of (D)OOH screens, using location data. Using location data, we only ask individuals located within a specific radius around a network of (D)OOH display screens.
​ID batch: By downloading a data file with IDs of exposed individuals.
Broadcasting framing: by downloading a data file.
