L’Oréal, the global leader in beauty and fragrance, debuted the YSL Beauty Tech Innovation Lab at the 2021 Dubai Expo. This innovative customer experience combines personality and brain data to recommend fragrances that participants are likely to enjoy. Rather than sampling each of the 27 YSL products, customers enjoy a quick, streamlined experience, answering a couple of questions about their emotional connection with fragrance and then sampling several “accords” (the base constituents contained in many perfumes) while wearing an EMOTIV neurotech headset. The result? Three fragrance recommendations that the customer is likely to enjoy and subsequently buy.
We live in an amazing era but with this often comes large demands on our time and multitudes of decisions to be made. Modern technology has promised us efficiency, but often the sheer number of available options can be overwhelming. Making our decisions easier by providing only relevant options is one of the ways technology is delivering on this promise.
How does it work?
At the heart of the experience is electroencephalography, or EEG. A wearable headset measures the electrical activity generated by your brain when your neurons fire. EEG is measured on the scalp and is an indicator of the underlying brain activity. For example, when you look at a human face, there is a very distinct electrical signature that can be measured with sensors over your visual cortex. Likewise, when you smell something you love or hate, recognizable electrical patterns can be observed. These changes in electrical activity can be recorded and then analyzed to provide insight into brain processes and emotional responses.
What can EEG tell us?
There are very few things that can be discerned from looking at “raw” EEG. Generally, scientists will look at EEG that is averaged across many people who are repeatedly exposed to a stimulus. EEG must be processed to clean up the “noise” and often transformed or otherwise mathematically altered in order to deduce any meaningful patterns. This is generally done long after the actual data collection has occurred.
However, one of the benefits of EEG compared to other brain activity measurements (like fMRI), is that it can detect processes happening very quickly. This makes it a tool uniquely positioned for “online” uses. In other words, producing insights in real-time rather than at a later time after all the data has been collected.
Producing real-time EEG insights
If EEG has to be cleaned and transformed to be meaningful, how can it be useful in real-time? By using previously-collected data, we can generate machine-learning algorithms that identify the general patterns of brain activity associated with specific circumstances. These algorithms can then be tuned to be sensitive to other situations. For example, we collected EEG from many people while they performed several frustrating tasks. We then developed a mathematical model of what this frustrated EEG looked like, and quantified the level of frustration between 0 and 100%. This model can be integrated into software that indicates when a new person is doing something that frustrates them.
EMOTIV has spent over a decade developing these real-time EEG insights. We have collected EEG from thousands of people under controlled experimental conditions. Using our unique database and methods, we developed several distinct machine-learning algorithms that can identify cognitive states in real time. These EMOTIV “detections” include: Frustration, Interest, Relaxation, Engagement, Excitement, Stress, and Attention. We fine-tune these detections and develop new ones through further testing.
Producing a customized, real-time fragrance detection
In the L’Oréal engagement, we leveraged our years of research expertise and experience developing sophisticated detection algorithms to devise and deliver a custom olfactory detection product. In partnership with L’Oréal, we collected data from hundreds of participants who were exposed to a range of YSL fragrances and accords. Participants also provided subjective ratings for each of these scents on various domains (likeability, familiarity, likelihood to buy, etc…).
Using the olfactory-related EEG, EMOTIV data scientists created a new detection algorithm that captures specific features related to each scent.The algorithm was trained to predict each individual’s self-reported “like/dislike” response to each exposure. Brain responses to a small set of accords are combined in a second model which predicts the likelihood a participant will find each of the 27 YSL fragrances appealing, then recommending the top 3 fragrances.
Integrating real-time detections into a consumer experience
In cooperation with L’Oréal, EMOTIV integrated our olfactory detection algorithm into a seamless app-based beauty experience. Alongside a beauty advisor, customers first answered eight emotional profiling questions that determined which six accords were most likely to stimulate them.
With six accords determined, the beauty advisor then fitted the EMOTIV headset to the customer and guided them through the exposure. Customers sampled each of the accords while brain responses were wirelessly transmitted to a mobile device. During this process, the EMOTIV real-time detections continuously adjusted the individual probability that the customer would find each of the YSL fragrances appealing.
The experience then provided customers with a personal profile. This likeability curve gave customers the opportunity to compare their profile with the YSL beauty community.
The experience culminated with three YSL fragrance recommendations. Two of these were predicted to perfectly align with the customer’s individual emotional and brain responses. To widen a customer’s sensory choice profile, the third recommendation was either a unisex or gender-opposite fragrance that still aligned with the predictive brain olfactory algorithm.
The resulting experience was a resounding success from the perspective of the customers, beauty advisors, L’Oréal, and EMOTIV. The “Scent-Sation” experience is due to roll out globally in YSL flagship stores later this year and through 2023.
The future of consumer-facing neuroscience
This partnership demonstrates how EMOTIV industry-leading hardware can be combined with our neuroscience processing pipelines to produce effective and accurate predictive applications.
With hardware and software technological innovations, we have entered an era in which neuroscience can be harnessed for a wide range of real-world applications.
Whether a company is looking to gain actionable insights into how customers are engaging with their products or seeking to bolster employee well-being through objective assessment of workflows and environments, modern neurotech is now positioned to be the go-to tool of the future. It is quickly supplanting subjective measures, such as surveys and questionnaires, as the most effective method for companies seeking to better understand their employees, their customers, their products, and their markets.
Find out more information about EMOTIV products here.