Understanding the Long-Term Use of Smart Speaker Assistants

Abstract

Over the past two years the Ubicomp vision of ambient voice assistants, in the form of smart speakers such as the Amazon Echo and Google Home, has been integrated into tens of millions of homes. However, the use of these systems over time in the home has not been studied in depth. We set out to understand exactly what users are doing with these devices over time through analyzing voice history logs of 65,499 interactions with existing Google Home devices from 88 diverse homes over an average of 110 days. We found that specific types of commands were made more often at particular times of day and that commands in some domains increased in length over time as participants tried out new ways to interact with their devices, yet exploration of new topics was low. Four distinct user groups also emerged based on using the device more or less during the day vs. in the evening or using particular categories. We conclude by comparing smart speaker use to a similar study of smartphone use and offer implications for the design of new smart speaker assistants and skills, highlighting specific areas where both manufacturers and skill providers can focus in this domain.

Privacy Attitudes of Smart Speaker Users

“…5 Based on the pilot, we made changes to the extension (see Section 3.1) and added two new questions. 6 We made no other changes to the survey, the two studies were conducted back-to-back, and the recruitment procedures were identical; as a result, the main study and pilot were substantially similar, so we report the results as one combined dataset.…”

“…To characterize the interactions our participants were reflecting on, we asked them several questions about the recordings they heard. (For more in-depth exploration of typical usage, see Bentley et al [6]) We first asked who was in the recording (Table 1). Over half of interactions were initiated by the respondent, with just under a third coming from other members of the household, including at least 6.75% that were attributed to children.…”

What Makes a Good Conversation?

“…This work shows that people tend to engage in limited and clearly delineated task-based conversations with IPAs. These include checking the weather, setting reminders, calling and messaging, playing music, launching other applications, information seeking, and interacting with Internet of Things (IoT) devices [2,20,35,38,54]. Although these devices promise much through their implied humanness [20], they fall short of the reflexive and adaptive interactivity that occurs in most human-human conversation [42].…”

A emergência dos smart speakers: desafios e oportunidades para o rádio no contexto do big data

“…Desta forma, observa-se um investimento na coleta de dados de maior dimensão, não necessariamente associados ao usuário em si, mas que revelam práticas amplas de consumo e a eficiência de estratégias de comunicação estabelecidas através dos dispositivos. Bentley et al (2018), embora não trabalhem especificamente com big data, analisam alguns dos rastros digitais deixados pelos usuários de smart speakers para tentar compreender o fenômeno de maneira mais ampla. Com a perspectiva de métodos digitais aplicados ao rádio (Lopez, Freire, 2018), os autores analisaram 65.499 interações realizadas durante 110 dias em 88 domicílios com assistentes de voz.…”