Japanese clothing retailer Uniqlo is trialling in-store “neuroscience” technology in Sydney and Melbourne that recommends t-shirts to customers according to their mood.
In trials planned over the month of October, customers can sit in an UMood both where they will be fitted with an off-the-shelf NeuroSky MindWave EEG headset.
After a minute where the headset is “calibrated” to the customer’s brainwaves, they are then shown a series of random short videos that are meant to provoke various brain and mood responses.
Those responses are fed into a proprietary system by Japanese company Dentsu ScienceJam.
The output from here is fed into a second algorithm, and finally the customer is presented with four different t-shirts they might like out of Uniqlo’s 600-plus collection.
As the customer is shown the four options, the brain response is measured instantaneously, and the t-shirts are ranked accordingly.
A final screen shows the most palatable t-shirt based on the customer’s brain activity – and Uniqlo presumably hopes they’ll trust this algorithmic salesperson enough to buy something.
The UMood booth is the work of Nuro, an agency that tries to work out the effectiveness of branding, advertising and marketing using neuroscience.
Nuro’s Dr Phil Harris – who is also an honorary fellow at the University of Melbourne – believes the technology has “a lot of potential”.
“When we give consumers a lot of choice, they have trouble making decisions,” he said.
“A tool which helps narrow the range can be really beneficial.”
While the UMood is naturally part customer gimmick, it is actually producing data that could be useful for a retailer like Uniqlo.
However, when Information Age quizzed Uniqlo Australia’s marketing director Tracey Lang, she said the company isn’t “approaching the trial that way”.
“I believe we can quantify the data,” she told Information Age.
“One of the things we’re interested in looking at is whether you might be able to look at what was Sydney’s [collective] mood vs Melbourne. That could be interesting.
“The other one that could be interesting is time of day, or is it sunny or raining and does that have an effect on people?
“But I’m not sure that we’ve thought too far on those things. We’ve just started.”
Depending on the type of data that UMood produces, one can also imagine the possibility for A-B testing: correlating the popularity of certain recommended items against what makes it to the point of sale, for example.
“You could probably do that,” Lang said – but again, it was not on the company’s radar at present.
One reason for this could be the temporary nature of UMood. Far from being an in-store fixture, it can only be tested October 10-11 in Pitt St Sydney, October 15-16 in Miranda, October 17-18 in Parramatta and October 24-25 at Emporium in Melbourne.
“We’re connecting it with our t-shirt collections now but we haven’t thought too far ahead on what we’re going to do with it after the trial ends,” Lang said.
“We also want to see how customers respond to it.”
I sat down in the UMood booth and had the EEG headset fitted by Uniqlo people dressed in white lab coats.
After a "baseline" reading that lasted about a minute, I watched a series of short video grabs: of a sinister dog's face, a kitten, a guy skateboarding, and VU meters going crazy.
I was given the mood "stormy" and recommended some t-shirts. They weren't too bad - at least they weren't all Disney characters as recommended for the previous guy, whose mood was read as "adventurous".