How to understand people's experience with voice-based agents, interfaces, and environments

What is it about?

Voice-based agents, interfaces, and environments are entering our lives in the form of voice assistants, smart speakers, conversational interfaces, and other "bodyless" but vocal machines. As we continue to take up these technologies, we must also understand how best to design them to fit our needs and expectations, as well as understand how they influence our attitudes and behaviour. We conducted a survey of the literature to see how researchers and practitioners are evaluating what we call "voice UX," or people's user experiences (UX) with these voice-based machines. We found that there was a great variety of ways, mostly relying on people's self-reports of their experience, and without much work conducted long-term or in people's real lives.

Featured Image

Photo by Ivan Bandura on Unsplash

Photo by Ivan Bandura on Unsplash

Why is it important?

Voice is an emerging key factor in the design of machines that interface with people. We need to make sure that we design machine voice carefully, with a view to people's needs and expectations. To test our designs, we need to evaluate them. Voice UX is a key way to do this. We need to standardize how we approach the design and evaluation of voice UX. We also need to explore more objective ways of understanding voice UX, such as through logs and metrics, observation, and controlled experiments. We need to go into the wild of people's lives and take a longer perspective on people's real experiences with these technologies in context.

Perspectives