Evaluating Usability of AI Based Voice Assistant Apps: A Study on Siri and Bixby

Voice assistants (VAs) are a rapidly evolving technology used for retrieving real-time information and controlling smart home devices. With the increasing use of smartphones and other smart devices, more people are becoming interested in voice assistants. However, unlike graphical user interfaces, voice user interfaces are more challenging to comprehend, thus making it difficult to evaluate their usability and user satisfaction. The objective of this research is to evaluate the usability of widely used AI -enhanced voice assistants and compare their usability performance. To achieve this, two voice assistant applications Apple's Siri and Samsung's Bixby were assessed. The study adopted Heuristic Evaluation and User Testing methods to systematically analyze their usability and performance across diverse tasks and user demographics. Results indicate that Siri outperforms Bixby in terms of speed and user satisfaction, particularly excelling in multi-turn interactions and error prevention. In contrast, Bixby demonstrated higher error rates and required more retries, especially when handling complex commands, although it performed better for basic commands. Additionally, the heuristic evaluation revealed that Siri adheres more closely to usability principles, such as system feedback, error prevention, and design simplicity, whereas Bixby struggles with flexibility and efficiency. As such this study enhances human-computer interaction by offering a comprehensive usability analysis of existing voice assistant apps that provides insights for improving voice user interfaces.