You're using the Facebook in-app browser. Please open the page in a normal browser to have the best experience.
x

FaceReader technology: What your facial expressions reveal while surfing the web

05/16/2018

by Nikolaus Franke

An innovative piece of software scans people's faces and analyzes the facial expressions they make to determine what they are feeling. Prof. Nikolaus Franke, Academic Director of the WU Executive Academy's Professional MBA Entrepreneurship & Innovation and Head of WU Vienna's Institute for Entrepreneurship & Innovation, has used this technology in the context of an international research project to gain a better understanding of why so many of the customers who take advantage of online configurators to design products according to their wishes give up after just 5 minutes. His findings are astonishing.

 

The glory and misery of product configurators

Thanks to the Internet and flexible production technologies, people can design a wide variety of products according to their wishes. Research evidence shows that customers feel this adds enormous subjective value to the things they buy. Also, they are prepared to pay up to 100% more for “their” products than for standard ones of the same objective quality. Online, countless configurators exist for products as diverse as clocks and watches, muesli, T-shirts, kitchens, cars and sofas (https://www.configurator-database.com/). While experts say “mass customization” is the future of marketing, the concept does not really catch on. Only a small percentage of those who use configurators actually complete the design process. Why is that? This question was the starting point for a research project I have carried out in collaboration with Franziska Metz (European Business School, Germany) and Page Moreau (Wisconsin, USA).

A woman is shopping online
Up to 100% more willingness to pay - but why is only a small percentage of those who use configurators completing the design process? Photo © CC0 Licence

How can you investigate what is going on inside the customer's head?

Finding out what people think and feel is difficult. You can, of course, ask them, and their answers may even be honest, though you never really know. Moreover, there is a problem if you want to get an idea of how their emotions change over time. You can hardly ask them repeatedly because doing so will distort the results. Hence, we were looking into possibilities of measuring the emotions of customers in such a way that the process does not influence their feelings. After carefully examining a number of options, we decided to tap into the potential of FaceReaders. This innovative technology scans a person's face and measures movements of the facial muscles by tracking changes in the positions of 500 key points without disturbing the individual in question. By applying empirically validated algorithms to the resulting record of facial expressions, you can analyze someone's current feelings based on the cross-cultural “basic emotions” suggested by Ekman (anger, sadness, disgust, fear, contempt, surprise and joy).

 

The investigation

We invited a total of 508 people to design a shoe with an online configurator. In an effort to reflect reality as closely as possible, we offered them the opportunity to buy “their” shoes at the end of the process. By means of a special auction procedure, we also collected data on how much they were willing to pay. Throughout the design process, we filmed the participants, recording their emotions and changes in their emotions during each and every second by means of the face-reading software. In addition, we collected other data that we linked with the FaceReader protocols at the level of the individual participant. Ultimately, we had millions of data items at our disposal, allowing us to analyze the emotions individuals experience in the course of designing products according to their wishes.

A man is laughing while using a tablet
Thanks to an elaborated face-reading software, emotions are measured. Photo © CC0 Licence

A roller coaster of emotions

Our findings show that, initially, customers have positive expectations of the design process. People love to be creative, and this is also true as far as the online world is concerned. However, their enthusiasm starts to fade once they face the challenge of familiarizing themselves with the configurator. “What? How is this supposed to work? Actually, what I was trying to do was … Ouch! Well, back to square one …” is what their facial expressions say.

 

  • During the first phase, customers are required to devote much of their attention to the design tool itself—and they become significantly less enthusiastic. After approximately 200 seconds, they understand how the software works and what they can do with it—only to run into another problem: “What do I actually want to achieve?” Researchers have known for quite some time that the preference insight of customers is low, i.e. they are not clear about their wishes and expectations regarding the design of products.

  • The second phase of the design process is marked by the frustration of not knowing what one wants. After 300 seconds, the emotions of customers hit a low. It is at this point in time that the risk of them giving up is particularly high. However, an idea is beginning to take shape—design activities are becoming more targeted in nature, and there is a sense of achievement.

  • You can easily tell that their mood is improving again. During this third phase, which, on average, lasts 200 seconds as well, customers get in the flow of things. They know how to operate the configurator, are clear about what they want and create designs that are getting better and better. Eventually, their emotions are even much more positive than they were in the beginning. The data we collected on customers' willingness to pay also show that they feel not giving up finally paid off for them.

 

What these findings mean for the design of configurators

The U-shaped emotional curve explains why so many of the customers who use online configurators give up. Apparently, they do not manage to overcome the low they experience 300 seconds into the process. These people turn away in frustration, not realizing that, later on, they would have felt increasingly positive emotions as a result of creating value for themselves. We can draw important conclusions from these findings as far as the design of online configurators is concerned. A variety of targeted measures help slow down the emotional decline, making it more likely that those interested in designing products according to their wishes actually complete the journey successfully.

A man is frustrated in front of his computer
When designing an online configurator it is important to overcome the users frustration after 5 minutes and give them a feeling of accomplishment. Photo © CC0 Licence

Using new technologies to gain a better understanding of customers

 

Our research project has shown that the software-aided analysis of facial expressions is a powerful tool when it comes to gaining a better understanding of the emotions of customers. This approach eliminates problems such as overtaxing, social desirability and strategic response behavior, which are typically encountered in the context of interviewing respondents. And, almost more importantly, the procedural nature of the data-recording process affords accurate insight into how emotions change over time. It makes good sense to use this technology also for other investigations, for instance into people's acceptance of new software versions, websites, product innovations or online advertising. From a business point of view, the emotions of customers are frequently predictors of crucially important behaviors such as their clicking habits or their decisions to buy.

 

To get more information about Entrepreneurship and the Professional MBA Entrepreneurship & Innovation click here.

Share this