Why and How do Neuromarketers Measure Brainwaves using EEG?
Why and how do neuromarketers measure brainwaves using EEG?
By Stephen Genco
EEG (short for electroencephalography) is one of the most popular methodologies employed by neuromarketers today. Here is a brief overview of how it works, what is measures, and why neuromarketers find it a powerful addition to their methodological toolkit.
How EEG Works
EEG measures brain activity by detecting and amplifying faint electrical signals, informally called brainwaves, that are emitted continuously by the brain. These electrical signals are the means by which our brain communicates and synchronizes activity across different anatomical regions. Variations in brainwave activity are indicators of changes in cognitive processing. Modern EEG equipment can take a snapshot of brainwave activity every 1-3 thousandth of a second, providing greater temporal resolution than any other measurement technology used in neuromarketing.
When many thousands of neurons fire together, they produce electrical potential or voltage differences across the scalp. German psychiatrist Hans Berger first reported in 1929 that these differences could be detected and measured with electrodes placed on the scalp and connected to a signal amplifier. EEG signal analysis has been a mainstay of brain research ever since.
What EEG Measures
There are three main types of EEG measurement that are regularly used in neuroscience research–brainwave frequency analysis, hemispheric asymmetry analysis (an application of frequency analysis), and event-related potential analysis. Neuromarketing today focuses mostly on only one of these, but the others have great potential for exploring consumer responses as well. They each emphasize a different aspect of the EEG signal, and each have their own pros and cons as a measurement technique for answering neuromarketing questions.
Brainwave signals naturally emitted by the brain have distinctive frequency characteristics. Electrical frequency is measured in cycles per second using the unit hertz (Hz). The most common frequency emitted by the brain is around 10 Hz, or 10 cycles per second. These frequencies change in response to different mental states, and also vary over time and across different parts of the brain. A convention has emerged to classify frequencies commonly associated with specific psychological processes into frequency bands named after Greek letters:
- Delta: less than 4 Hz, dominant frequency in dreamless sleep
- Theta: 4-8 Hz, associated with internally focused processing, such as memory activation and conscious concentration
- Alpha: 8-12 Hz, the brain’s “default” frequency, dominant when the brain is in a relaxed state and suppressed under the influence of attention
- Beta: 13-30 Hz, associated with alertness, active attention, and reward expectation
- Gamma: greater than 30 Hz, associated with information processing, learning, and emotional processing
Two metrics are commonly used to measure brainwave frequencies: power measures the amount of brainwave activity occurring at a particular frequency over a period of time; coherence measures the consistency or correlation of brainwave frequencies across different parts of the brain. Greater power means greater activity or energy in a given area at a given frequency; greater coherence between regions often means the regions are communicating as part of a cognitive process.
An application of EEG frequency analysis that has become especially popular in neuromarketing is the measurement of frequency band asymmetries (differences) between the left and right frontal regions of the brain. Neuroscience researchers discovered many years ago that these asymmetries were associated with approach and avoidance motivation with regard to an object of attention (for a review, see Harmon-Jones et al., 2010). When experiencing approach motivation, people generate more beta and gamma band power and less alpha band power in the left frontal hemisphere compared to the right, and the opposite is true when experiencing avoidance motivation.
Event-related potentials (ERPs)
One subset of EEG measurement is the analysis of event-related potentials (ERPs). As the name implies, ERPs measure brain signals that occur directly in response to some event. By averaging responses to a large number of very rapidly presented exposures to stimuli, ERP researchers can isolate the sequence of brainwave responses–called components–elicited by the stimulus. By comparing ERP components for different stimuli, inferences can be made about different types of nonconscious and conscious responses, such as personal relevance, allocation of attention, expectancy violation, and emotional judgment.
ERP components are named based on their polarity (positive or negative) and their latency (how long after stimulus exposure they occur). For example, the P300 component is a positive potential that occurs at approximately 300 milliseconds after exposure to a stimulus. It is sensitive to personal relevance and shifts in attention. The N400 component is a negative potential that occurs approximately 400 milliseconds after exposure to a stimulus that is incongruent either semantically or with a person’s beliefs or knowledge. The LLP or Late Positive Potential is an ERP component that occurs at approximately 600 milliseconds after exposure and has been linked to emotional judgment and valence change. Many other ERP components have been identified, each representing a different event-related cognitive response (Luck and Kappenman, 2011).
How EEG Is Used in Neuromarketing
Frontal hemispheric asymmetry is used by many neuromarketers as a measure of approach-avoidance responses to products and brands. Frontal asymmetries have been found to reflect immediate, preconscious motivational responses to marketing stimuli like media ads, brands, and physical products. Variations in frontal asymmetries occurring as quickly as 200 milliseconds after exposure to a marketing stimulus have been found in some studies to be potential predictors of subsequent consumer choice and behavior, sometimes predicting responses better than consumers’ own expressions of their wants or preferences (Ravaja et al., 2013; Ramsøy et al., 2018).
Brainwave frequency analysis has a long tradition in market research, dating back to the 1970s (Krugman, 1971). Early efforts to associate responses to marketing stimuli with overall power in different frequency bands lacked precision, but more recent efforts have found intriguing correlations between beta-band power and product preferences and gamma-band power and marketplace success (Boksem and Smidts, 2015). More recently, sophisticated statistical techniques have been applied to summarize power and coherence patterns throughout the brain; these approaches have begun to yield interesting predictions of consumer behavior at both individual and market performance levels (Ariely and Berns, 2010). One promising application has been to examine “intersubject synchrony” of brainwave activity while viewing entertainment or advertising materials. Recent studies have found this metric to be predictive of both immediate engagement and subsequent marketplace success (Barnett and Cerf, 2017; Dmochowski et al., 2014; Christoforou et al., 2017).
Event-related potentials (ERPs) have been used relatively sparingly in neuromarketing research. Some have speculated that this is because they are difficult to explain to nonscientists, but they are also more demanding of researchers, requiring more data collection, more precise experimental designs, and more sophisticated analytical methods to arrive at conclusions. These hurdles should not obscure the significant potential of ERP studies for neuromarketing, however. For example, the P300 ERP component has been used to assess brand associations, such as whether a new brand extension fits in with existing brand expectations (Ma et al., 2008), and the N200 component has been used to predict product purchase preferences (Telpaz et al., 2015).
Limitations of EEG
EEG has certain limitations that researchers and buyers of neuromarketing services should keep in mind:
- Designing, running, and interpreting the results of EEG studies requires PhD-level expertise. Despite some commercial claims to the contrary, EEG is not a do-it-yourself methodology.
- Metrics can be challenging to understand and difficult to interpret, unlike more direct behavioral measures such as facial expressions or eye-tracking fixations.
- ERP studies require repeated measures to separate signals from the background noise of unrelated brain activity (the “signal-to-noise” problem). This can make it difficult to measure responses to novel stimuli such as new products or packaging.
- EEG is not a suitable technique for measuring electrical activity originating deep within the brain, such as in the emotional and memory centers, because those signals become too faint and dispersed before they reach the surface of the scalp.
The Bottom Line
In summary, there are many reasons why EEG continues to be an extremely popular measurement technology in neuromarketing. It is the only brain measurement technique that can capture brain activity at the speed of cognition. It measures brain activity directly, rather than indirectly, through behaviors and choices. In addition, it benefits from a vibrant research and technological innovation community. In recent years, EEG equipment has become more inexpensive, portable, and wireless, opening up new possibilities for mobile, in-store, and virtual reality studies. New statistical and machine-learning techniques are starting to be used to decode and interpret the EEG signal at the level of the full brain, portending many new and original findings that would have been unobtainable only a few years ago (Halchenko and Hanke, 2010).
Barnett, Samuel B., and Moran Cerf. "A ticket for your thoughts: Method for predicting content recall and sales using neural similarity of moviegoers." Journal of Consumer Research 44.1 (2017): 160-181.
Boksem, Maarten AS, and Ale Smidts. "Brain responses to movie trailers predict individual preferences for movies and their population-wide commercial success." Journal of Marketing Research 52.4 (2015): 482-492.
Christoforou, Christoforos, et al. "Your Brain on the Movies: A Computational Approach for Predicting Box-office Performance from Viewer’s Brain Responses to Movie Trailers." Frontiers in neuroinformatics 11 (2017): 72.
Dmochowski, Jacek P., et al. "Audience preferences are predicted by temporal reliability of neural processing." Nature communications 5 (2014): 4567.
Halchenko, Yaroslav O., and Michael Hanke. "Advancing Neuroimaging Research with Predictive Multivariate Pattern Analysis (MVPA)." (2010).
Harmon-Jones, Eddie, et al. "The role of asymmetric frontal cortical activity in emotion-related phenomena: A review and update." Biological psychology 84.3 (2010): 451-462.
Krugman, Herbert E. "Brain wave measures of media involvement." Journal of Advertising Research 11.1 (1971): 3-9.
Luck, Steven J., and Emily S. Kappenman, eds. The Oxford handbook of event-related potential components. Oxford university press, 2011.
Ma, Qingguo, et al. "P300 and categorization in brand extension." Neuroscience letters 431.1 (2008): 57-61.
Ramsøy, Thomas Z., et al. "Frontal Brain Asymmetry and Willingness to Pay." Frontiers in neuroscience 12 (2018): 138.
Ravaja, Niklas, et al. "Predicting purchase decision: The role of hemispheric asymmetry over the frontal cortex." Journal of Neuroscience, Psychology, and Economics 6.1 (2013): 1.
Telpaz, Ariel, et al. "Using EEG to predict consumers' future choices." Journal of Marketing Research 52.4 (2015): 511-529.