hero image

Researchers from Penn State conducted a study on customizing voice assistants like Alexa or Siri to sound similar to the user and found that people have a strong preference for voice-activated devices that sound extroverted, speaking faster, louder, and in a low pitch. This customization makes the voice assistants more likable.

Customized voce assistants receive better ratings

According to the study, when voice assistants are customized to match a user’s personality, people tend to give better ratings to the device. Feedback suggests that the customized voice assistants sound more socially and intellectually attractive, making them more trustworthy in the eyes of users.

Study co-author S. Shyam Sundar stated that the tendency to associate perceived resemblance with credibility was particularly noticeable among individuals who personalized their interaction by selecting a favored voice for the virtual assistant.

According to the study, users were less receptive to hearing COVID-19 vaccine misinformation when the voice assistant sounded similar to them. For instance, only 38% of unvaccinated individuals became more open to vaccination when exposed to vaccine misinformation shared by a virtual assistant. The authors of the study were surprised by this resistance to certain information from the voice assistant.

Eugene Snyder, an assistant professor at the New Jersey Institute of Technology, explains that people often resist persuasive attempts from information sources such as social media influencers or pundits. He adds that when unvaccinated individuals were confronted with misinformation from a voice assistant that sounded similar to them, they exhibited resistance. However, more research is required to understand this reaction fully, as the unvaccinated participants constituted only 27% of the sample size.

Customizing VA improves interaction

The study, involving 400 participants, explores human information processing. Participants self-rated extroversion and were randomly assigned to groups experiencing different voice assistants: robotic, customized, or personalized. Control group received introverted/extroverted robot voices. The findings offer insights into how individuals respond to various voice assistant voices, enhancing understanding.

Customizing voice assistants resulted in improved interactions with the assistants, particularly among individuals whose voice assistants shared a similar sound to their own. The findings indicate that incorporating personalization and customization features can enhance user experiences.