DeePay: deep learning decodes EEG to predict consumer’s willingness to pay for neuromarketing

Abstract

There is an increasing demand within consumer-neuroscience (or neuromarketing) for objective neural measures to quantify consumers’ subjective valuations and predict responses to marketing campaigns. However, the properties of EEG raise difficulties for these aims: small datasets, high dimensionality, elaborate manual feature extraction, intrinsic noise, and between-subject variations. We aimed to overcome these limitations by combining unique techniques of Deep Learning Networks (DLNs), while providing interpretable results for neuroscientific and decision-making insight. In this study, we developed a DLN to predict subjects’ willingness to pay (WTP) based on their EEG data. In each trial, 213 subjects observed a product’s image, from 72 possible products, and then reported their WTP for the product. The DLN employed EEG recordings from product observation to predict the corresponding reported WTP values. Our results showed 0.276 test root-mean-square-error and 75.09% test accuracy in predicting high vs. low WTP, surpassing other models and a manual feature extraction approach. Network visualizations provided the predictive frequencies of neural activity, their scalp distributions, and critical timepoints, shedding light on the neural mechanisms involved with evaluation. In conclusion, we show that DLNs may be the superior method to perform EEG-based predictions, to the benefit of decision-making researchers and marketing practitioners alike.

A systematic review on EEG-based neuromarketing: recent trends and analyzing techniques

“…The articles that deal with identifying emotion through brain data using different stimuli, such as e-commerce product images [ 8 , 36 ], video ads [ 37 , 38 ], product shapes [ 39 ] etc., have been considered under this cluster. These articles mainly focused on EEG-based emotion identification/preference prediction by adopting different ML techniques to train the models to predict the future preferences of the consumers.…”

“…DL models are best suited for predicting outcomes when there is a large dataset to learn from. Hakim et al [ 36 ], Teo et al [ 39 ], Aldayel et al [ 40 , 122 ], Al-Nafjan [ 47 ], Göker [ 51 ], Georgiadis et al [ 160 ], and Alimardani and Kaba [ 161 ] used different DL models for the preference prediction tasks. Using DL models, Hakim et al [ 36 ], Teo et al [ 39 ], and Göker [ 51 ] achieved an average classification accuracy of 75.09% [ 36 ], 79.76% [ 39 ], and 96.83% [ 51 ], respectively.…”

“…Hakim et al [ 36 ], Teo et al [ 39 ], Aldayel et al [ 40 , 122 ], Al-Nafjan [ 47 ], Göker [ 51 ], Georgiadis et al [ 160 ], and Alimardani and Kaba [ 161 ] used different DL models for the preference prediction tasks. Using DL models, Hakim et al [ 36 ], Teo et al [ 39 ], and Göker [ 51 ] achieved an average classification accuracy of 75.09% [ 36 ], 79.76% [ 39 ], and 96.83% [ 51 ], respectively. Using different feature selection algorithms along with the DL classifier Al-Nafjan [ 47 ] could achieve an average classification accuracy of 99%.…”