Brand Lift's Biggest Mistake : Why ad recall falls short

Introduction

Consider this a Brand Lift intervention. At a recent marketing industry event, we asked everyone who came to our stand a couple of quick questions. The first one was, "Which advertising exposure measurement technique is most reliable?" A. Passive exposure measurement or B. Ad Recall based measurement. An eye-watering 68% said Ad Recall. These were people from agencies, AdTechs and media owners; brand managers, Insights Directors and CEOs. People who should know better. While shocking, it's perhaps unsurprising. Brand lift measurement is complex—it's often misunderstood, mis-sold or undervalued. So we wanted to set the record straight.

Brand Lift measurement using Ad Recall relies on a simple question: "Do you remember seeing this ad?". But memory isn't simple; it's imperfect, influenced, and often wrong. People forget ads they did see (false negatives), recall ads they didn't see (false positives), and mistake one brand's ad for another (misattribution). If your Exposed and Control groups are unreliable, your entire measurement is flawed. Ad Recall should not be used as a proxy for exposure. It's a useful datapoint within the marketing funnel, not the basis of reliable measurement.

People recall ads they didn't see, forget ads they did, and confuse one brand for another. When your Exposed and Control groups are built on these shaky foundations, your Brand Lift measurement is fundamentally flawed.

Methodology matters because accuracy matters

We looked at the results of over 2,000 brand lift studies in our digital benchmark database to put Ad Recall to the test. By comparing the answers to our Ad Recall question with the passive exposures we measured (using high-confidence pixel tagging), we were able to quantify the extent of the inaccuracy from false negatives and false positives. The data showed that, in 8 out of 10 cases, Ad Recall was wrong. Relying on participants' memory introduces both over- and under-reporting. From the outset, the results are flawed: groups are misattributed through human error, contaminated by bias, and blind to subconscious effects. On this basis, true Brand Lift simply cannot be measured.

Understanding customer bias

When we analyzed the extent of selection bias from customers, the problem was clear. Looking at 15 passive measurement studies, we found that Ad Recall for customers was on average 3.1 times higher than for non-customers. Customers were consistently more than three times as likely to say they recall seeing an ad, despite both groups having comparable exposure to the campaigns. This means that in an Ad Recall based Exposed group, 7 out of 10 people are likely to be existing customers (weighted for sample size). This is not measuring Brand Lift, it's measuring an existing relationship with the brand and artificially inflating your campaign results. It's as bonkers as asking Tesco customers if they've heard of Tesco.

It's a well-known fact that familiarity with a brand primes you to notice and remember its advertising. Even if you balance for brand usage, Ad Recall magnifies three problems:

• Selection bias - Those who recall an ad are often already more engaged with the brand or category, more attentive to advertising in general, or further along in the purchase process. Higher brand consideration among recallers may reflect these traits, not campaign impact.
• Behavioural bias - Asking respondents whether they remember an ad primes them to think about the brand, inflating subsequent responses on measures like consideration or intent. This can manufacture "lift" that isn't truly there.
• Subconscious effects - Ads work in many ways outside conscious awareness. Relying on recall captures only immediate, conscious effects and ignores deeper, longer-term brand building.

Customers and prospects will answer Ad Recall questions differently. Relying on Ad Recall as the starting point for your Brand Lift measurement makes it harder to reveal the levers to grow.

Comparing the approaches

Ad Recall Based Measurement: 

1. Participants are asked if they recall seeing a specific ad or brand within a recent timeframe. Their responses separate them into two groups: exposed (people who say they've seen the ad) and control (people who say they haven't seen the ad).
2. Memory is messy and error-prone. Some will think they saw the ad when they didn't (false positives), and others won't remember seeing it when they did (false negatives). Due to these errors, the exposed and control groups will be contaminated with participants who should be in opposing groups.
3. The methodology estimates the impact of ad exposure by comparing the two groups, but when the groups consist of a mix of truly exposed and unexposed participants, the measured impact will be entirely distorted.

Passive Exposure Measurement: 

1. We embed a lightweight 1x1 pixel into the ad creative to passively measure ad exposure. When the ad is served, the pixel allows us to build a real-time exposure graph, automatically creating exposed and control groups.
2. As exposure is tracked directly through pixel tagging, we know with a high degree of confidence who has and has not seen the ad. Participants are then placed into exposed and control groups.
3. The data is objective and unaffected by human error, ensuring clean, uncontaminated sample groups.

By comparing responses between the exposed and control groups, we can isolate the true impact of the campaign. Passive measurement avoids the biases and flaws of recall-based methods, giving you Brand Lift metrics you can trust for decision making.
 

The cost of cheap Brand Lift

Our industry is full of smart people. So, given its shortcomings, why is Ad Recall still being used? The most common answer is a lack of understanding, especially at a senior level. An Ad Recall based approach is often considered "good enough" when the value of Brand Lift measurement is not fully understood. In organizations where Brand Lift data is used strategically, as an input into future media planning and models, the importance of accuracy is obvious. In our experience, an Ad Recall approach creeps in when Brand Lift is little more than a box tick at the end of a campaign and the insight isn't being used for future optimization.

Which leads us on to cost. Quality Brand Lift measurement is complex and Ad Recall based studies are easier and cheaper to run , making them the fallback option when budgets or campaign scale are limited. This is a false economy. At best the results will be giving you a rough read on creative impact. You won't be getting metrics that correlate with real-world outcomes. If you're continually seeing positive Brand Lift results with no visible impact on business performance then trust will gradually be eroded and confidence in your brand investment will be lost.

In short, Ad Recall should never be the basis for serious Brand Lift measurement. Data quality matters and with Brand Lift you get what you pay for. If you need data you can trust for media investment decisions, passive methodologies are by far the most reliable solution.