- Retailers in Asia have been piloting a solution from tech company Wayvee that can identify heightened emotional arousal in shoppers as they review products and their pricing
- The system works with electronic shelf labels (ESL) for dynamic pricing changes based on the data, as well as tracking analytics about store layout and promotions
How are shoppers feeling when they enter a store, browse through products on the shelves, and face the listed prices? While understanding the feelings of shoppers could have value for retailers, the only way to access their thoughts would be asking the shoppers their opinions or analyzing facial expressions.
One technology company is using an AI-based store analytics system that uses a different approach—it captures frequency-modulated continuous wave (FMCW) transmission responses around shoppers, and can evaluate whether their emotions are aroused based on their micromovements.
Wayvee developed its solution to offer emotion analysis for brick-and-mortar retail with what co-founder and CPO Alex Ovcharov calls a new method for stores to identify price and display issues by analyzing customer emotional responses.
The camera-free, AI-powered, in-store analytics technology provides real-time insights into shelf reactions. That means retailers could resolve problems up to 10 times faster than relying solely on sales reports—optimizing dynamic pricing strategies based on demand, and enhance the shopping experience and drive sales, Ovcharov said.
Dynamic Pricing Based on Emotions
The Wayvee smart shelf solution detects shoppers’ emotional responses by analyzing changes in micromovements including heart rate and breathing rate, thereby understanding how the experience in the store is being received by the subject. The Wayvee sensor is used to better understand purchase intent and desire to buy a product. By knowing how shoppers respond to products and prices, stores can then dynamically adjust their Electronic Shelf Labels (ESL) pricing.
In one example, a retailer in Singapore completed a pilot program in its grocery store. The prices were linked to the Wayvee smart shelf solution. The system estimated emotional responses to selected products and the company could make changes to its pricing accordingly. The retailer reported an 11 percent increase in drink shelf sales after using the system.
“Unlike traditional, in-store analytics methods that rely on cameras or facial recognition, Wayvee offers a privacy-respectful approach using radio waves to analyze customer emotional response through micro-movements, such as breathing and subtle body shifts,” said Ovcharov.
History in Gaming
With fellow cofounder Anton Timashev, Ovcharov established the business known as Sensemitter to provide emotional recognition of gaming participants, during gameplay as well as when viewing advertising. Based on facial coding and AI, the system uses a gaming device’s existing camera to detect users’ reactions and deliver a second by second analysis of emotions to the game publishing studios. The studios then use that data to improve their games.
Sensemitter officials began next to look at other industries that might benefit from such analysis. The team landed on consumer product goods (CPGs) and retailers, and Wayvee resulted.
Such systems could help with analyzing the effectiveness of the advertising and the packaging design for the products, according to the CPG’s and retailers Wayvee spoke with. However, these companies indicated they didn’t want to use a system as part of a study tool where volunteers are recruited from panels to look at a design and provide their responses, and they didn’t want to deploy camera-based systems into their stores.
FMCW for Micro Movement Detection
Ovcharov, whose background is in neuroscience, discovered how radio waves could be used to detect human movement at the most minute level. He found that such systems are capable of detecting changes in breathing and heart rate.
He pointed to a level or response that shoppers can exhibit as they look at products, which could be neutral, or could be positive or negative. Known as the arousal-valence model, the technology assesses changes in breathing and heart rate. The technology goal is to identify purchase intent based on a shopper’s emotional response to what they are seeing.
Ovcharov said the most accurate parameter is arousal, “as it is the key to action and arousal in front of the shelf is the key to purchase.”
Such data could be collected with cameras. However, “technically a camera needs to have a very consistent view of the face,” said Ovcharov so the store environment doesn’t lend itself to this kind of system due to people continuously moving in the store as well as privacy issues. By using radio waves instead of cameras, the solution offers anonymity for the shoppers. The system has no way to detect who it is standing in front of the product. But it does know their intent based on arousal, for instance.
How it Works
Wayvee provides sensors that can be mounted near key product displays—typically products that generate the largest revenue, or that are especially popular with shoppers. It can be connected to the cloud-based software via Wi-Fi and powered at a local outlet.
The sensor transmits short-wave, electromagnetic 60 GHz wave signals. The signal reaches a shopper and bounces back to the sensor. This takes place in high volume and very fast. AI-based analytics of that response can then determine how long someone is standing there and how much they are moving.
The movement changes can be highly subtle, even a change in pulse and breathing affects the transmission response.
Someone would typically need to stop in front of a display and its sensor for about two seconds before their emotional arousal could be measured.
Integration with Pricing
Retailers using the technology thus far already have some level of advanced store management systems like electronic shelf labels, shelf replenishment or AI display planning, said Ovcharov, “so they all actually can change their stores environment in real time.”
For example, changes of pricing on electronic shelf labels can be done remotely, in real time, by store management or a retailer’s office. They can decrease prices if people appear to be dissatisfied with their price. Wayvee can send an alert to the task management system in the store so that the employees will go check the display to see what the problem may be.
Additionally, retailers could use the data with support from customers or loyalty members to better identify what is taking place at the shelf level. Wayvee is developing an integration with crowdsourcing platforms that will guide crowd members directly to shelves with issues, thereby providing another tool related to store audits.
Pilots Began Last Summer
CJ Express, a large retail chain in Thailand has been piloting the technology in five of its stores. AI enables the measurement of arousal, indicating a purchase intent. When purchase intent doesn’t lead to a purchase, it’s a sale lost due to high price. “The only barrier to purchase for someone with intent is the price,” Ovcharov said.
Other pilots have been underway for the past six months. In most cases, said Ovcharov, sensors are mounted according to most spontaneously purchased products including drinks, alcohol, snacks and sweets. A sneaker store chain is testing the system to evaluate shopper response to a pair of shoes.
“We are currently in talks with several big fashion clients they are very data conscious,” he added. The data can be correlated with other issues such as display arrangements, out-of-stocks or the environment, such as music in the area.
The company is moving its solutions to the U.S. market now. In fact, Wayvee has headquarters is in New York with two offices in Europe: in Cyprus for software development and in Poland for hardware development.