The Body Knows: Using Neuroscience and AI To Predict Advertising Success

Go back to the last video ad that caught your eye. Was it the creative that stopped you from scrolling? Was it the message? Or perhaps it was the way it made you feel? While you might know why you liked it, your brain and body were telling a different story through subtle signals you weren't even aware of - small expressions on your face and tiny changes in your skin's sweat response.

We still use focus groups and surveys to tap into what consumers think they like about an advertisement. But new research shows there's a more reliable, faster, and cheaper way to predict advertising success: measuring the unconscious signals your brain and body send through, like facial expressions and skin conductance (or emotional response through the skin - think of ‘skin crawling’ when you’re disgusted about something you see).

Using artificial intelligence to analyze these reactions, researchers can now predict with 81% accuracy which ads will resonate with consumers before a campaign launches.

If you have relied on traditional testing methods in marketing so far, you might want to consider this new approach. It’s a practical way to test and optimize your advertising: you measure viewers' unconscious physical responses rather than just their words.  It’s a more reliable predictor of which ads they’ll prefer than their conscious opinions. 

A New Way To Understand Ad Effectiveness Before Campaign Launch

Neuroscience has proven to be a valuable tool for understanding buyer behavior (Eijlers et al., 2020; Plassmann et al., 2015). It allows brands to grasp how consumers process information and make decisions based on neurophysiological responses.

Researchers from China, Portugal and Spain tested how powering neuroscience insights with AI can open a new door to marketing strategy and advertising effectiveness.

A new experiment was set out to prove that it’s possible to anticipate the type of ads consumers prefer, so useful before launching or for optimizing existing campaigns. Participants watched six video ads for different cosmetics brands, while their body responses were recorded using two scientific methods: Facial Expression Analysis (FEA) and Electrodermal Activity (EDA).

FEA uses computer vision technology to analyze changes in facial expressions. Along with attention and engagement levels, seven basic emotional reactions were tracked in real time (Joy, Anger, Fear, Surprise, Sadness, Disgust, and Contempt) (Ekman, 1992).

EDA was used to monitor subtle changes in skin sweat activity during ad viewing. Why is this useful? Because it gauges the intensity of the emotional response, rather than the specific type of emotion (positive or negative). Fluctuations in emotional arousal levels can indicate comfort or discomfort. It makes it possible to predict consumer preferences for products or brands during advertising exposure. 

Machine Learning for Ad Prediction

This study creates a precedent by the addition of an analytical layer using a machine learning framework. The research team applied three different machine-learning approaches to analyze the data collected from facial expressions and skin response measurements. The Random Forest (RF) technique achieved an 81% rate in accurately analyzing patterns in emotional responses. To understand which factors influenced the predictions the most, Explainable AI (XAI) techniques were implemented to help marketers trust and effectively use the results. The combination of research methodologies avoided complexity in analysis and streamlined the interpretation of neuroscience results. Future studies in neuromarketing could benefit from applying this technique, in order to enhance understanding of consumer behavior.

Key Factors of Successful Ads

So what makes an ad successful? The results showed four key elements that drive consumer preferences: attention and engagement, expressions of joy, and the absence of disgust. One of the most significant findings was the strong correlation between joy and engagement, discovered across five out of six tested advertisements. This suggests that when viewers experience joy, they're more likely to remain engaged with the content.

Practical Applications for Marketers

When testing and optimizing ads, you don’t need to rely only on what consumers say. Measuring their physical responses can highlight factors particularly valuable during the creative process, where even small adjustments can significantly alter a campaign's success. Use a small test group to predict your ad's effectiveness before launch, by measuring unconscious responses.

Focus on creating positive emotional connections rather than just rational arguments. Ads should evoke more joy while minimizing disgust to drive engagement. The connection between joy and engagement suggests creating more positive content. Although it's not just about happy ads. It’s equally important to minimize elements that trigger disgust. 

The Accessibility of FEA and EDA

Both EDA and FEA are more affordable than complex brain imaging tools. They are relatively simple to implement in typical market research settings. The real-time nature of the feedback is ideal for agile marketing where creative teams can quickly iterate and improve their work based on scientific data rather than subjective opinions.

The combination of physiological measurements and artificial intelligence analysis allows for a practical way to test different versions of ads before the full launch. Specific moments that drive engagement can be identified during testing. The ad content can then be optimized based on emotional responses. It is a new way to reduce the uncertainty in advertising effectiveness. After all, we’re dealing with human nature. 

Limitations and Future Potential

The findings are promising, but they can’t be generalised. The study focused specifically on cosmetics advertising, using a small female Chinese sample. 

At the same time, it presents possibilities for better understanding and predicting advertising effectiveness. That means more engaging ads that resonate with the audience at a deeper, emotional level.