measures user reaction to ads and music via neuromarketing. This is how it works

The Ukrainian company is working on a device that will allow reading reactions and engagement of users in advertising, games, and their experience on online store sites. This practice is called neuromarketing, and, globally speaking, it is nothing new; however, in Ukraine, according to CEO and co-founder of the company Vitaliy Miniaylo, his team is the first to try to do this. AIN.UA visited the company’s office, tested the device, and tells how it works.

How did the idea come about

Globally, many companies use neuromarketing, that is, the assessment of feelings and user engagement in a product using their physiological data. For example, the well-known game development corporation Valve (Counter-Strike, Dota 2, Left 4 Dead) works with OpenBCI and Tobii to read the emotions of players who are testing games. It is done using eye-tracking and other biometric data (you can read more about the scientific side of neuromarketing research here).

According to the company’s co-founder, the idea came about when the team was looking for how to accurately measure the effectiveness of the interface when developing a website or application for a customer.

“We’ve heard about neuromarketing before, but in a rather abstract way. At the same time, we had and still have a strong R&D team with extensive experience in content promotion. For example, our team has been making YouTube videos trending since the dawn of the platform. Then we decided to try neuromarketing: it turned out that almost no one in Ukraine does it. We began to piece together hardware, software, and so on to combine all this into a single complex,” says Vitaliy Miniaylo, co-founder of the IT company

It was supposed to be a product that encompasses neurophysiology, marketing, and neural networks. The company’s co-founder is now writing a thesis on the theoretical and methodological foundations of creating artificial intelligence systems at the Kyiv Polytechnic Institute. The team has eight developers in the field of neural networks and ML.

Co-founder of the company Vitaliy Miniaylo. Hereinafter all photos by Olha Zakrevska / AIN.UA

Scientists in the field of neurophysiology also cooperate with the company: Sergii Tukaev, a senior researcher at the Institute of Biology and Medicine of the National Taras Shevchenko University of Kyiv, Sergii Danilov, a candidate of biological sciences and an EEG and iitracking expert, and Viktoria Kravchenko, a candidate of biological sciences.

How it works

The team has assembled a working device in the form of a “smart helmet,” with various sensors and software, with the help of which it conducts neuromarketing research, tests different hypotheses, and fulfills commercial orders.

The complex of devices by is a set of equipment that includes various sensors, a helmet, and eye-tracking devices (the set may vary depending on the research). The sensors and helmet are put on the user’s head and arm, and the user watches the necessary ads, music videos, or the site interface.

The device includes:

  • an electroencephalograph that reads the electrical activity of the brain: with its help, data is taken by which it is possible to understand the level of involvement and attention of the user, as well as, with some approximation, the emotions that they are experiencing;
  • an eye tracker or a camera that reads the user’s eye movement, the direction, and fixation of the gaze: with the help of this data, you can see how the user’s gaze moved across the picture, where it stopped and lingered;
  • device for measuring heart rate;
  • assessment of facial expressions (this tool is used less often than others);
  • they also measure the galvanic skin response (bioelectric activity of the skin to measure the work of the sweat glands).

Together with programs for processing this data, it all works this way: the user gets seated in a chair, then the team puts on sensors and shows the user, for example, a music video or an advertising video. At the same time, the team monitors EEG indicators, measures the pulse, and monitors the pupils’ dilation and eye movements. That is how the first raw data is received, which is processed and cleaned of noise. Scientists-neurophysiologists are working on the data, and the output is data based on which conclusions can already be drawn.

EEG data show how much a person was involved in viewing at a particular moment, what emotions they experienced. Data from the eye tracker: where the person was looking and where their gaze lingered. Also, such a camera can measure the diameter of the pupil. The team learns to classify emotions, track eye movements, etc., more accurately using neural networks.

Data processing

To access the raw data, the team uses proprietary software (connected via API). Based on mathematical models and machine learning, this raw data processes and predicts emotions, interest, and other metrics. The company is still working on the accuracy of the forecast and assessment.

For example, working with EEG and eye tracker data:

  • removal of artifacts and noise;
  • analysis of the spectral and spatial characteristics of the signal;
  • depending on the frequency, strength and place of brain activation, conclusions are drawn about the cause of the signal (the team determines the degree of interest, involvement, etc.).

How it looks

Here is an example of how EEG data is taken for neuromarketing research (video from the company’s YouTube). It is raw data that requires processing to then present the result of the study to the customer in an understandable language. So far, neurophysiologists are working with EEG data “manually”, but over time the company expects to automate this process.

Here’s how iTracking works: The red lines and circles show the direction of the eye movement. The bigger the circle, the longer the user has been looking at the same point.

The editor of AIN.UA tested the device on herself. She did not manage to record EEG because of the hairstyle, but the device recorded eye movement, focusing of the gaze at certain points, as well as the pulse and the galvanic reaction of the skin. The latter reaction shows how the amount of sweat on the surface of the skin changes and can serve as additional data to clarify the feeling a person is experiencing.

In the future, the price of such an “out-of-the-box” device could be several thousand dollars.

Who are the clients

Vitaliy Miniaylo says that such data, unlike, for example, polls and focus groups, gives access to the true reactions and feelings of people who will watch an advertisement, listen to a music video, or play a game. This data can be interesting for marketing agencies, online stores, video, music, and game production studios.

He gave the example of several cases of using such a device. For example, it was used to analyze how listeners reacted to new songs by a famous artist. And this is an example of the fact that the reactions that users voice and what they actually experience can be different.

“We made a ranking of songs by the famous Ukrainian singer. For example, when users shared their reactions to the song with dirty words in its title, some of them wrote that they didn’t like it, probably thinking that it’s inappropriate for a song to have such a name. But at the same time, we saw from the EEG that their brains responded positively to the song, suggesting they actually liked it,” says Minyaylo.

The company has also told us about a case with one of the major Ukrainian banks which ordered neuromarketing research. The task was to assess how readers perceive the bank’s banner ads on several news websites and identify the elements attracting the most attention. The research used eye-tracking followed by interviews with respondents to gain a better understanding of their experience. They said that the style of the banners was nice, and the ads were clearly visible on the website. At the same time, only 10% of respondents could remember who the advertiser was. Eye-tracking helped mark out the following reasons:

  1. Too much superfluous explanatory information on the banners.
  2. The focus on the single word “free” and mixing of the name of that free service
    with the main text.
  3. The use of redundant visual objects – the buttons “Order” and “Open.”
  4. Improper accentuation: the human face photo occupied ⅓ of the entire image,
    while the information on the banners was about transaction bonuses.

The bank was recommended to:

  1. Remove faces and redundant buttons “Order” and “Open” from the banners.
  2. Position the information about the service itself in the foreground, with a note that it is free, below.
  3. Use words and images that the audience will associate with a bank of choice.

When those amendments were applied, the bank’s recognizeability increased, as well as the reaction on the advertised service. The overall information reading rate for one of the banners grew from 27% to 58%.

The complex for neuromarketing research is not the only service that the company is working on. Some other projects planned to be developed by are a cashier-free shop with eye-tracking cameras; a device assessing how many people actually looked on a billboard and what emotions they had while watching, and other ideas.

The company estimates that neuromarketing will become common practice over time. “Our idea for this solution is to come in one box and, over time, become part of creators’ culture for generating the content,” the founder explains.