WIRED Retail is our annual exploration of the ever-changing world of commerce, featuring leading technologists, entrepreneurs and creatives innovating in sectors as diverse as robotics, virtual reality and the future of home delivery. For all our coverage from the event, head over to our WIRED Retail hub.
Companies spend billions trying to find out what consumers think. But now, deep learning, computer vision and cognitive neuroscience could unlock the mystery for a fraction of the cost. "Since the beginning of time we've failed to understand how people feel," Ken Denman, president and CEO of engagement and emotion analysis firm Emotient told the audience at WIRED Retail. "The reality is, we've just been guessing."
Denman claimed a rapid rise in computational power and data analytics techniques are finally allowing us to detect, in real-time, the micro-expressions that give away what people are really thinking. "Its difficult to get people to tell us what they really feel. And that's what you want to know. That's what your business depends on," explained Denman.
In one example he showed Emotient's emotion and engagement engine gathering data from people taking a survey on Carl's Jr. Super Bowl advert.
The technology, which Denman said was 95 percent accurate, was able to recognise joy and anger, spotting the key emotional triggers by identifying tiny fluctuations in facial expressions. The online survey was completed by 300 people in four days and cost $6,000 (£4,000). What did the data show? That men enjoyed the advert and women didn't. "There are powerful, powerful insights here," said Denman. The technology can also be used to analyse crowds, spotting who is looking at big-screen advertising at a sporting event. The software can also detect the race, age and gender of who is looking. Once you know who's looking and when, Denman argued, you can sell advertising based on in-depth knowledge.
"With cameras everywhere, and cameras on all the time theatre's a huge opportunity," he said. But this also raises privacy concerns. When cameras everywhere are able to understand our emotions, what does that mean for our freedom?
Denman explained that all the pixels they capture are 'exploded' when processed, making it impossible for anyone to put them back together. "We work on an anonymised, aggregated basis," he said.
could also be placed in shops and by billboard advertising to see how people are responding to advertising. "Think of this as next generation focus groups," he said.
This article was originally published by WIRED UK
