How are brands tracking emotions to improve their content?

31 : 07 : 2017 Digital : Technology : Media
Emotion measurements for The Revenant by Lightwave Emotion measurements for The Revenant by Lightwave

Last week, the team at Disney Research explained at the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) that they are using artificial intelligence (AI) to assess how much their audiences are enjoying their films, minute by minute.

They claimed that the technology is so effective that, after analysing a film-goer’s face for 10 of those minutes, it can then predict that person’s future emotional state throughout the rest of the film. From that, it could then alter the film’s overall narrative in real time. The research uses what Disney calls ‘factorised variational encoders’ – this sophisticated new technology generated 16m data points from 3,179 audience members for the report presented at CVPR.

‘We are all awash with data,’ says Markus Gross, vice-president of research at Disney Research, ‘so it is critical to find techniques that discover patterns automatically. Our research shows that deep learning techniques, which use neural networks and have revolutionised the field of artificial intelligence, are effective at reducing data while capturing its hidden patterns.’

We originally highlighted the role emotions play for brands in our macrotrend The E-motional Economy, where we urged brands not only to speak in the varied language of emotions, but also to look into the potential of using facial-recognition technology to tailor their content accordingly. And although the idea of having the film you’ve paid to watch now watch you just as intently might seem slightly unsettling, it is something that brands are increasingly taking note of.

In March of this year, Mars teamed up with emotion measurement technology specialist Realeyes – a company which has also worked with Heineken, LG, Audi and BMW, among others – to measure how people were feeling when watching its adverts. When that data was compared with sales data from each of the ads included, Realeyes found that emotion-tracking was able to correctly identify the impact of each ad on sales 75% of the time; in other words, to identify which ads sell and which ads do not.

CBInsights recently reported on Facebook’s newest patent, granted in May 2017. The patented tech will monitor users’ typing speed to predict their emotional state and push advertisements or content to correlate with however they are feeling. It follows a similar patented tech that the company filed in 2014 which looked to read facial expressions through a phone or laptop’s front-facing camera.

Condé Nast is also looking to emotional exploration to better its sponsored content. While the company is not attempting to read emotional activity in real time, it recently partnered with Neuro-Insight, which specialises in neuroscience-based market research, to monitor brainwave activity among readers and measure the impact of its content on memory encoding and emotional intensity.

As emotion tracking becomes more accessible and brands become hungrier to see how their content is making their customers feel, it will be interesting to see how these innovations will make their way into the mainstream without eroding consumer trust. Disney, the world’s second-largest media conglomerate and Millennials’ most intimate brand, will have to work especially hard to maintain trust among customers who know that Disney films, theme parks and products might be monitoring their every mood.

For more on why brands should care about their consumers’ emotional lives and offer content that spans a full spectrum of feelings, see our macrotrend The E-motional Economy.