Emotion AI: changing the face of UX? (Part 2)

In part 1 of our article, we introduced emotion and the increasing presence of emotion AI. But what are the benefits and barriers to this to this new technology?

Benefits of emotion AI

There are several benefits to emotion AI, which all link back to providing an optimum experience for the user. We have summarised our top five below:

1. It produces richer research insights

This is obviously our number one! We always encourage our clients to view usability testing sessions either in person or via a live stream so that they can hear rich, impactful feedback about their product or service directly from the user. Now, imagine if we could tap into the user’s emotions directly. Combine it with another methodology, such as eye tracking, and now we know exactly where a user is looking on a webpage and how their emotion changes as a result.

Improved understanding of user emotions at each interaction during testing would enable businesses to optimise the visual and UX design of products or services in order to maximise their impact. Identifying the emotions of participants at specific touchpoints, would also arm researchers with more data to probe around in order to uncover deeper insights. This information could add significant power to reporting within businesses to drive product improvements.

2. Allows for real-time personalisation

Imagine if you were feeling angry on the train journey home, and your Instagram filtered out negative posts, whilst Spotify suggested switching to a music playlist that it knew made you smile. By monitoring emotion in real-time, products or services have the power to adapt to meet user needs at any specific moment.

Several large corporations including Amazon, Google and IBM already have patents for technology that personalises content and advertisements based on the user’s current emotional state. Although this may initially come across as manipulative marketing, it also presents several benefits to the user such as reducing workload and promoting wellbeing.

3. Creates a more natural human machine interaction

Automated conversational devices such as smart speakers and chatbots now commonly replace humans during information-seeking and customer service journeys. Emotion AI has the potential to reduce caution around this type of interaction, by enabling a more natural ‘humanised’ conversation, where emotions are considered in decision-making algorithms. For example, based on a user’s emotion, a machine could choose to respond in a different tone of voice or to route the user through an alternative conversation flow. This has the potential to improve a design, as functionality is optimised for the user’s current emotional state.

For example, it was recently disclosed that Alexa (Amazon’s AI assistant) will soon be able to speak in different styles to convey six different emotions. Soon, if Alexa hears that you are upset, she could show greater empathy, by using a softer voice and suggesting ways to help. In this way, the product would build a stronger relationship with the user, as it begins to mirror human emotion and behaviour.

4. Promotes health and wellbeing

There are numerous research studies showing that poor regulation of negative emotions can increase the risk of health conditions. There are two ways in which a product or service can negate negative emotion; by removing pain points in the user experience or by providing a real-time response as negative emotion occurs. To do the first, the answer is simple – research!

By considering a representative sample of users during product development, problems can be resolved before they impact on a larger population. The second option returns to emotion AI, which has been programmed to trigger a response. For example, if a vehicle detects that the user is stressed, it could change the ambient lighting colour to blue, as research suggests that blue is effective in reducing stress levels. Of course, what works for one user, may not work for another, and this response could be made more effective through machine learning and personalisation.

5. Increases customer satisfaction and loyalty

A good user experience produces a happy, loyal customer. Used in the right way, emotion AI has the potential to empower products and services to be flexible, understanding and above all, positive. By making the user feel valued as an individual, brands can increase the chance of positive experiences being shared with others, thus boosting their customer base.

Barriers to emotion AI

As with all new technology, there are several barriers standing in the way of emotion AI being integrated into products and services.

1. Privacy is taken one step further

In order to measure emotion accurately, devices will need to collect data such as heart rate and facial expressions from the user. Although many people already allow devices such as smart watches to capture basic biometrics, the intrinsic nature of emotions takes a step further towards personal boundaries. You only need to look at recent media coverage to understand that many humans are afraid of having their data stolen or used inappropriately. As well as being open about the data being collected, and providing the user with full control over permissions, products will need to present a good value proposition to users. What is it going to do so well that is justifies sharing something that users already try to hide from other human beings? I’ll take a guess that targeted advertisements won’t cut it.

2. The reliability of detection technology

Although the emotion detection market is growing, there are still questions around the technology and the extent to which it can be applied across the world. With most solutions relying on facial recognition, they read individual facial-muscle movements, such as the raising of an eyebrow, and piece them together to infer a particular emotion. However, a recent study suggests that the facial communication of emotion varies substantially across cultures, contexts, and even by individuals within the same context. It seems you can’t escape individual differences.

Where products plan to provide a response to emotions, a false reading could result in worsening rather than improving of the user experience. With a growing expectation that technology should work straight away, a negative user experience is likely to reduce users’ openness to the idea.

3. It’s not quite here!

Despite advances in emotion AI, we are yet to see its capability fully utilised in a consumer product or service. Sorry, but you’ll have to keep reviewing products manually for now.

At System Concepts, we believe that with the correct methodologies and expert facilitation, it is still possible to uncover genuine emotion.

If you are wondering what is going through your user’s minds as they discover, use and reflect on your product, then please get in touch. Our experienced consultants will be more than happy to help.