Summary of My Role
I led a multi-method comparative study of Google Home and Alexa voice assistants, examining both general usage patterns and branded app experiences. Over several months, I conducted lab testing with 50 users, managed a diary study, and developed new frameworks for analyzing voice interactions.
Research Objectives:
Understanding Platform Differences Compare how users interact differently with Google Home and Alexa, focusing on both successes and failures in natural conversation.
Analyzing Voice Interaction Patterns Document how users naturally speak to voice assistants and how they adapt their language when interactions fail.
Evaluating Branded Experiences Understand what makes branded voice apps (like Domino's) succeed or fail, and develop guidelines for voice interaction design.
Research Activities
- Lab Testing (50 users) Conducted comparative usability tests where users attempted similar tasks on both platforms, including both general commands and branded app interactions.
- Diary Study (2-3 weeks) Users recorded their daily interactions with their assigned voice assistant, capturing both successful interactions and points of friction.
- Utterance Analysis Performed detailed analysis of user speech patterns, command structures, and error likelihood across both platforms.
Key Contributions
Hierarchy of Needs for Voice Interaction
Developed a framework showing how basic usability needs must be met before users can appreciate more sophisticated features. For example, if users can't reliably open an app, they won't care about its engaging personality or brand-specific features.
Natural Speech Patterns
Our analysis revealed that users don't naturally adapt their speech when voice commands fail. Instead, they tend to repeat the same phrase with increasing frustration, suggesting a need for better feedback and guidance from voice assistants.
Branded Experience Guidelines
Created a reusable framework for evaluating branded voice apps that went beyond simple command success rates. This helped clients understand how their brand voice could be maintained while still meeting basic user needs for clarity and efficiency.
Analysis Tools
Developed new methods for analyzing voice interactions that combined traditional usability metrics with analysis of success of different sentence structures. These tools helped bridge the gap between user research and voice interface development.
Want to know more? Shoot me an email at uxnoah@gmail.com, or schedule a time with me here.