Skip to Content

Preventing Customers from Getting Stressed Out

Measuring people’s physical reactions helps companies improve the user experience.

The restaurant chain Boston Market had a problem: it couldn’t figure out why people weren’t returning even though they said they liked the food. When customers were surveyed, they themselves couldn’t articulate why they didn’t come back.

Stressful shopping: Readouts from the Q Sensor show a customer’s skin conductance in green; the peaks indicate higher levels of physiological arousal, which often indicate stress. (More information on next page.)

Then Boston Market hired Shopper Sciences, a consulting company that tried to answer the question a different way. In addition to asking dozens of customers for their opinions, it also asked them to wear a portable stress sensor while they ate. The wrist sensors, which are made by a startup called Affectiva, detect galvanic skin response—how conductive, or “clammy,” the skin is. This measurement usually correlates with levels of physiological arousal, either positive excitement or negative stress.

Shopper Sciences found that diners were stressed out by eating at Boston Market. “The old Boston Market served food out of metal trays, and you were expected to eat chicken with a plastic fork,” says John Ross, CEO of Shopper Sciences. “The collective gestalt was terrible.” Ross says the findings helped inspire a redesign at the chain, which now offers food on real plates, with metal knives, forks, and spoons. “It’s the same food quality and price point,” he says, “but now it’s being delivered in a new way.”

With the availability of smaller, cheaper, and more powerful sensors, businesses are finding new opportunities to collect data on how their customers feel about them. Researchers have long used galvanic skin response to glean insight into people’s emotions, but gathering this data required volunteers to sit still with wires strapped to their fingers or palms. Affectiva’s device, called the Q Sensor, measures galvanic skin response many times per second while a subject goes about normal activities. The sensor also measures the wearer’s movements and temperature. Later the device can be plugged into a computer so the information can be analyzed.

“At the highest level, emotions drive decisions,” says Dave Berman, the CEO of Affectiva. “A lot of times people think they know what they feel but can’t articulate it, or aren’t even aware of it. We provide tools that help people better understand how their customers are feeling.”

Disney is evaluating the Q Sensor and other companies’ products as tools for measuring viewers’ reactions to TV shows, particularly on ABC and ESPN, says Duane Varan, the chief research officer at Disney’s media and advertising lab. “It’s a very powerful set of measures,” Varan says. However, he cautions that data from these physiological sensors could easily be misread: a high degree of arousal could mean someone liked a scene rather than having been annoyed by it.

To address such potential misreadings, Affectiva has released a facial recognition program to accompany the Q Sensor; the software can make use of video observations to help determine whether a person is reacting positively or negatively.

Stressful shopping: Readouts from the Q Sensor show a customer’s skin conductance in green; the peaks indicate higher levels of physiological arousal, which often indicate stress. The middle green line shows skin temperature, and the three colored lines at the bottom reflect the customer’s wrist movements. Shoppers who called themselves “casual” makeup users tended to get far more stressed about matching the colors of two different kinds of cosmetics (chart at bottom) than “frequent” makeup users (top chart).

When the consumer electronics website CNet wanted to test the effectiveness of a mobile app it was developing, it turned to Shopper Sciences, which put Q Sensors on several dozen customers at Best Buy. The customers, who had downloaded the app while it was in beta and agreed to the test before coming to the store, were looking for cameras, camcorders, computers, and other electronics. The app let them scan bar codes to call up product reviews, rankings, and social-media tools on CNet.

“You could see shopper stress drop immediately when they turned to the mobile application to do research on the product,” says Ross. His team also found that the shoppers were more likely to click on banner ads in the mobile app when the ads contained information about the item they were searching for. After the customers had made their purchases, Shopper Sciences followed up with interviews and questionnaires. “We were able to link what they talked about and what we observed and also how their body reacted,” Ross says. “The intersection of those three things gives you a really rich understanding of what’s going on.” The takeaway from the CNet experiment: it’s stressful to decide whether a big purchase is worthwhile. “We are figuring out ways to lower that [stress] so it’s easier to make a decision and then it’s easier to get more people to buy,” he says.

Similarly, a cosmetics consulting company wanted to better understand the behavior of customers shopping for makeup in supermarkets and drugstores. Shopper Sciences recruited around 150 shoppers, put stress sensors on them, and observed them by cameras or in person. Ross says the sensors revealed that casual makeup wearers tended to be confident in picking the shade of lipstick they liked, for example, but that their stress “dramatically increased” when they tried to match one cosmetic to another, like lipstick to eye shadow. Now the company is developing a mobile app that lets shoppers test combinations of products that might work with their skin type.

Keep Reading

Most Popular

The inside story of how ChatGPT was built from the people who made it

Exclusive conversations that take us behind the scenes of a cultural phenomenon.

Sam Altman invested $180 million into a company trying to delay death

Can anti-aging breakthroughs add 10 healthy years to the human life span? The CEO of OpenAI is paying to find out.

ChatGPT is about to revolutionize the economy. We need to decide what that looks like.

New large language models will transform many jobs. Whether they will lead to widespread prosperity or not is up to us.

GPT-4 is bigger and better than ChatGPT—but OpenAI won’t say why

We got a first look at the much-anticipated big new language model from OpenAI. But this time how it works is even more deeply under wraps.

Stay connected

Illustration by Rose Wong

Get the latest updates from
MIT Technology Review

Discover special offers, top stories, upcoming events, and more.

Thank you for submitting your email!

Explore more newsletters

It looks like something went wrong.

We’re having trouble saving your preferences. Try refreshing this page and updating them one more time. If you continue to get this message, reach out to us at with a list of newsletters you’d like to receive.