I stumbled upon a ground-breaking piece of research by emlyon about Voice User Interaction (aka VUI) which redefines the role a voice activation in applications design. Whereas it is flavour-of-the-month to claim that voice is going to replace other user interfaces and namely touch, Margherita Pagani’s AIM Centre of research for artificial intelligence in value creation proved that both types of interfaces have different implications with regard to the way that users interact with brands, engage with them, and feel about them.

VUI – Voice User Interfaces — as used in smart speakers, but not only — is often touted as a miracle cure to user interaction issues. However promising this technology might be, and it is, there are no miracles and touch-based interaction should not be overlooked.

Whereas it is clear that voice interaction has a very promising future for the design of applications, be it on smart speakers or other devices which are using voice recognition systems, it is not true to state that voice will replace other interaction methods, at least in the very near future.

As Margherita Pagani points out, there is a possibility that users are not feeling very comfortable with voice interaction right now, but things may evolve in the future. As always with innovation, and despite what people think, patience is a key ingredient.

I interviewed Margherita last week to know more about the research, which will soon be published by the Journal of interactive marketing, and of which I will give you a few glimpses in this piece.

The role of VUI in consumer interaction with brands

Voice interaction (VUI) and brand interaction with users

Margherita Pagani – emlyon

Interview of emlyon’s Margherita Pagani: Margherita Pagani is a Professor of digital marketing at emlyon business school. She is also the academic co-director of the Master of Science in Digital Marketing and Data Science, taught in Paris and Shanghai. She is also the Director of the AIM research centre on artificial intelligence and value creation.

Emlyon’s centre of Artificial Intelligence is looking at how new technologies and artificial intelligence improve user interaction. We study how technology can create new experiences and their impacts on marketing outcomes. One of the research areas we focus on at the research centre is the role of VUI (voice user interface) and touch in the interaction with the end user.

In this study, we compared the roles of modes of sensory perception, voice and touch and the two combined together, when brands interact with users. What motivated us to pursue this research was that more and more companies are now looking at new technologies to develop improved experiences of their interaction with the user.

Thus, voice interaction is getting more and more popular. It could be a more natural way to communicate with the customer.

Our assumption at the beginning was that the more means of sensory perception a brand uses to interact with its customers, the more the customers develop a natural relationship with the brand.

We also found that there weren’t a lot of articles or studies in the academic environment that address the role of voice.

You have carried out the research using two different experiments, one with what you call a “hedonic product” and one or more “utilitarian product”

We considered two different settings. For the first, we had a “hedonic product”; hedonic means something that is not fundamentally useful or doesn’t satisfy a basic need. We took a specific type of perfume from Sephora (LVMH household name for worldwide personal care and beauty stores) and then compared it with the same experience for a utilitarian product, namely a French frozen food supplier Picard.

Our goal was to see how technology impacts a user’s trust in a brand and his/her final relationship with it. The novelty of this study was that we looked at the interaction mediated by an interface and its effect on the brand.

So we created two groups of participants for this experiment. One group had to interact with the brand using technology encompassing both touch and voice, while the other had to use touch only. Then, we asked them to complete a set of tasks.

At the end of the process, they had to make phone calls to the companies Sephora and Picard, so as to ask for some information to give the experiment a real context. Finally, after their interaction using the interface on their mobile phones, we asked them to rate, on a scale of 1 to 7, the different experiences they had and their levels of trust.

At the end of the day, is voice activation (VUI) leading to more engagement from users or not?

What we found was unexpected compared to our assumption at the beginning. Eventually, we found that if we use only touch, it’s more impactful and creates more engagement with the brand.

We expected that if we use voice plus touch, i.e., more means of sensory perception, the experience would be more natural. The underlying assumption was that if we use both touch and voice, the brain would be more stimulated with these modes of sensory perception, and we’ll have a more natural relationship with the brand.

A sample of the research protocol used by M. Pagani et al.

But, the results were the opposite; it actually creates less engagement for different experiences with the brand. We explain these results with the help of the “dual coding” theory.

This theory states that when you have a means of sensory perception, you have to activate different parts of your brain. An increased number of such means implies more cognitive effort. Of course, this should be contextualized as it is true only when the technology is completely new. So, if you have to interact using a technology that is completely new, the brain has to make more cognitive efforts, because you have to understand the technology first and then develop a relationship with the brand. That’s perhaps the best way to explain these unexpected results.

Does that mean what we call natural language processing is not really natural?

It refers to a more natural experience than natural language. If we use voice, the interface of our relationship with the brand is more natural. It showcases a more real-life relationship as we are “talking with the brand”.

Nowadays, brands are trying to develop and facilitate more interaction using different modes of reality perception.

Does that mean eventually that maybe there’s too much hype about voice interaction (VUI), and that touch-based interaction still has a long way to go?

Yes, exactly. It could also be the familiarity with technology and the fact that voice, in the beginning, doesn’t work properly, and so you have to work in an iterative manner. This necessitates other cognitive efforts. Interestingly, this result is also confirmed by another, more robust research that is in progress at the research centre.

It’s very similar to what happens when a customer interacts with robots for the first time. Since she is not familiar with this interface, she exercises the same cognitive efforts. This will reduce the impact in terms of engagement and the experiences that a brand can leave.

A view of the theoretical model used in the research (sneak preview of the forthcoming article to be published in the journal of interactive marketing)

What is your recommendation to industrialists and marketers who want to develop applications, should they go for a voice interaction or touch-based or both?

Voice is very promising, though the brands are considering also on how to design applications using touch and voice at the same time. Maybe the result that we are looking at now in this study and in other studies about the interface, is that in the initial phase of adoption of new technologies, it’s important to put emphasis on the usability of the interface. This is because if the interface doesn’t work properly, it will have a negative impact on the brand.

Only when a piece of technology is more mature and the customer is more familiar with it, a brand should build up a marketing strategy to leverage more on its benefits for creating new experiences. Therefore, a suggestion for companies is to look at voice, it is very important.

At the same time, they shouldn’t underestimate, particularly in the initial phase, the difficulties that could be created by technology with regard to the usability of its interface.