“Mediaprobe specializes in quantifying audience engagement with TV and CTV content and advertising. We’re introducing a new signal for media quality to the market, which is highly sensitive to the second-by-second engagement levels of audiences with the content they’re watching,” explains Pedro Almeida, CEO of Mediaprobe. “This signal is derived from measuring Galvanic Skin Responses (GSR) at scale, capturing physiological reactions from consumers as they watch content in their homes.”
ALAN WOLK (AW): Can you explain what “galvanic skin response” actually is, what you are measuring and how do advertisers then use those metrics to their advantage?
PEDRO ALMEIDA (PA): In a nutshell, Galvanic Skin Response (GSR) measures the activation of the sympathetic nervous system. It detects changes in the activity of sweat glands in the hands or fingers, serving as a robust, validated, and highly sensitive indicator of emotional activation. For example, our team conducted a meta-analysis of all scientific studies examining the relationship between physiological responses and emotion in videos. We demonstrated that GSR is the most sensitive indicator of perceived emotional intensity in video content.
We’ve developed the technology to collect GSR data from panels of thousands of participants, sync it in real time with the TV content they’re viewing, and process the data through a pipeline that feeds into a platform for visualizing results.
We measure GSR at scale by distributing thousands of proprietary, highly accurate wireless GSR sensors to our panels, allowing participants to watch TV while being monitored. The resulting data is used in various ways: assessing the value of programs as a context for advertising (e.g., does this program or ad break position deliver an engaged audience?), optimizing how ad breaks are structured, activated, and valued, and analyzing the contextual relationship between program content and ad creatives to enhance emotional engagement with ads.
AW: TV measurement has long focused on basic viewership—how many people saw an ad and, occasionally, the actions they took afterward. Was there a particular moment or experience that made you realize emotions could be a more powerful way to build on that foundation to better understand ad effectiveness?
PA: The idea that traditional viewership data is foundational but needs to be complemented by other engagement metrics isn’t new. Engagement acts as a bridge between exposure and the subsequent behavioral response to ads. This falls into the realm of “attention” metrics, though we believe that term doesn’t adequately capture what we aim to measure for TV.
We’ve identified three key principles for an effective metric:
It must be highly sensitive to how much the audience is engaged with the show.
It should carry over from content to advertising (i.e., the level of engagement with content should predict engagement with ads).
It needs to correlate meaningfully with external outcomes.
The challenge is finding a metric that adheres to these principles and serves as a meaningful measure of media quality. Audience emotional engagement with TV content, as measured through GSR, fulfills these criteria.
This gives us a powerful signal of how much viewers are engaged with the content, and it carries over into the advertising break and commercial activation. When we think about media quality metrics for TV, this signal is precisely what we need: it indicates audience engagement with content and its impact on advertising effectiveness. In short, we believe this is the best measure for what the industry truly means when referring to “attention” in the context of TV.
AW: Can you share examples of how brands have used your insights to improve campaign performance?
PA: The majority of applications for our metrics occur when publishers aim to demonstrate the quality of their programming as an ideal setting for ads in order to optimize the value of each ad impression within a break or commercial integration. We help determine how to structure ad breaks, decide where to insert them during programs, and identify key ad slot positions to assign value.
On the buy side, we’re increasingly conducting what we call 360º campaign tracking. Brands use our technology and data throughout the entire campaign lifecycle—from pre-testing to assess whether an ad can elicit an emotional response, to designing the optimal media plan for maximum audience impact, to in-flight tracking of the campaign. Once the campaign is live, we monitor the creative units to ensure they’re performing as expected. These distinct use cases both rely on our metrics to disentangle the effects of creative and media on the emotional response to advertising, helping clients make more informed decisions.
For instance, we’ve had clients track their creative units in-flight and decide to pull underperforming ads after seeing they weren’t resonating emotionally with target audiences. Even though these ads had been pre-tested and approved using their standard tools, our tracking revealed a lack of emotional engagement in one or more key demographics. Based on these insights, clients removed the underperforming creative and then increased the frequency of a better-performing one.
We’ve also seen clients revise sponsorship plans after discovering that certain properties they were sponsoring didn’t generate meaningful emotional engagement with audiences. On the sell side, we’ve worked with clients to completely transform how they structure and market their ad breaks.
For example, we collaborated with a client on one of their key tentpole programs to pinpoint, second by second, the optimal placement of ad breaks in order to maximize emotional impact. By optimizing break placement, we achieved significantly higher emotional engagement for each ad via carry-over from the content to the break, making each ad more memorable.
AW: How do you see the introduction of emotional metrics influencing future ad models, like shoppable TV or interactive content?
PA: Producing an emotional response to an ad relies on having the right creative and ad format, as well as being in the right content at the right time. Figuring out ad timing within the content is essential to reaching viewers during moments of high engagement. One exciting way our metrics are being utilized for advanced advertising is through their second-by-second resolution, which allows us to pinpoint moments when audiences are maximally engaged and activate commercial activity around those moments.
We’ve been testing protocols to help clients determine exactly when to insert ad breaks or activate sponsors—for example, using a lower third or a billboard during key moments of a broadcast. With these metrics, clients can evaluate whether the timing of their activations truly aligns with peak audience engagement.
On the measurement side, we help assess whether specific ad formats can impact audiences as intended. Increasingly, we’re being used to activate contextually aligned ads and to determine whether this alignment generates a significant emotional response from viewers.
Today, the industry can contextually align ads on numerous dimensions—for example, aligning an ad with the relevant product category, the mood of the content, the emotions displayed in surrounding scenes, the overarching theme of the show, or the theme of the ad itself. The question becomes: which type of alignment actually produces a significant emotional response in the audience?
We’re increasingly being called upon to identify which contextual dimensions are truly effective, calculating the resulting emotional lift without requiring a full attribution study, which can be time-consuming and resource-intensive.
AW: Are there specific types of content (e.g., live sports or news) where emotional measurement is more valuable, e.g. understanding the delta between the emotions generated from the program and those generated from the ad?
PA: Actually this type of measurement is relevant for every type of content. What’s essential is a large database that shows how different genres perform in terms of generating reliable emotional responses across various audiences.
Let me give you some examples. News and sports have very distinct emotional profiles, or morphologies. This is intuitive if you think about it, but it’s crucial to measure. In sports, for instance—while different sports vary slightly—there’s typically a consistent pattern:
High emotional intensity before the match,
Long stretches of low emotional intensity,
Interrupted by peaks of very high emotion when something significant happens.
This pattern of low emotion punctuated by unpredictable peaks of high emotion is characteristic of sports, though it’s more pronounced in some sports than others.
When we look at news, we typically observe a steady pattern of moderate to high emotional engagement throughout the program, though this depends on the nature of the news being reported. By contrast, well-crafted scripted shows have a distinct and standardized emotional morphology. For example, successful, binge-worthy programs often end episodes with measurable cliffhangers, motivating viewers to watch the next episode.
Understanding the emotional morphology of each genre and format is critical when optimizing the placement of advertising and sponsorship activations. It helps ensure that ad timing and context align with audience engagement patterns for maximum impact.
AW: What is the one thing you wish more people in the industry understood about emotional engagement metrics?
PA: This may seem trivial, but I wish more people understood that “emotion”—or even “attention,” for that matter—is just a word. The same word can represent a variety of methods to operationalize it, while different words can sometimes describe the same concept.
For instance, when people talk about emotions, they often mean very different things. Some might refer to discrete basic emotions like happiness, disgust, sadness, or fear. Others might focus on affective dimensions, such as whether someone is activated and whether that activation is pleasant or unpleasant. Still, others might think about moral emotions, like shame or guilt. These are all different interpretations of the word “emotion.”
So when vendors claim they have a new measure of emotion, it’s essential to look under the hood and understand exactly what is being measured and how. As I mentioned earlier, our use of “emotional engagement” refers to a specific methodology. What we’re actually measuring is the galvanic skin response of audiences, synchronized with TV content. This provides a robust signal of how engaged people are with the content, and this engagement carries over into advertising.
There’s nothing magical or inherently special about the word “emotion.” The same applies to “attention.” These are just terms for phenomena that can be measured in different ways—some methods are more useful in certain cases than others.
AW: What’s next for Mediaprobe - what are you most focused on in the year ahead?
PA: We’re focused on significantly growing our commercial footprint and expanding our data acquisition capabilities. On the publisher side, we’re seeing rapid adoption of our metrics as a way to evaluate media space and optimize the timing and placement of commercial activity.
On the buy side, we’re noticing increasing interest from media agencies and brands in having an actionable and scalable measure of engagement with television. This has been a key area of growth for us as well.
We’re also expanding our data acquisition efforts by not only growing our own panels but partnering with third-party panels to achieve the next level of scalability.
Additionally, we’re investing heavily in AI to model our data, steadily moving toward predictive models of engagement across different shows and genres. These predictive capabilities will enable broader adoption of our metrics for planning, forecasting, and not just measurement.
Finally, we’re deeply committed to advancing the media measurement industry, with a particular emphasis on quality measurement. We believe 2025 will be a pivotal year for driving progress in this space.
