Profiling the AI speaker user: Machine learning insights into consumer adoption patterns

Area of Science:

• Consumer Behavior
• Artificial Intelligence (AI)
• Internet of Things (IoT)

Background:

• The rapid evolution of media technology necessitates understanding consumer adoption of AI speakers.
• Effective advertising and marketing strategies require accurate identification of potential AI speaker consumers.
• Previous research has not fully profiled emerging IoT consumer segments.

Purpose of the Study:

• To identify characteristics of current AI speaker users.
• To predict potential future consumers of AI speakers.
• To inform targeted advertising and marketing strategies for AI speaker adoption.

Main Methods:

• Utilized machine learning classification techniques including decision trees, random forests, support vector machines, artificial neural networks, and XGboost.
• Analyzed data from the 2019 Media & Consumer Research survey (N = 3,922) from the Korea Broadcasting and Advertising Corporation.
• Developed and validated predictive models to profile potential AI speaker consumers.

Main Results:

• The XGboost model demonstrated superior performance in predicting consumer profiles.
• Key potential consumers identified are individuals aged 45-50 and 60-65.
• These individuals are characterized by social media activity, varied content preferences, weekday internet use, weekend cable TV viewership, and 5G technology awareness.

Conclusions:

• Advanced machine learning effectively profiles potential AI speaker consumers beyond simple prediction.
• Media consumption habits, lifestyle patterns, and technological understanding are critical factors in AI speaker adoption.
• These findings provide actionable insights for developing focused and effective marketing campaigns for IoT devices.