"UTAUT-Based Mobile Financial Services Churn Prediction Dataset"

"This dataset consists of a well-structured set of survey answers intended for analyzing adoption behavior and churn dynamics among users of mobile financial services (MFS) operating in the Sylhet and Moulvibazar districts of Bangladesh. The survey was conducted in January-February 2026. The dataset contains 938 responses collected with a mixed approach to data gathering, including online tools, social media, and field work in remote locations.The data include 24 variables describing the demographic profile (residence, literacy, and income level) of MFS users, their transactional behavior (frequency and length of inactivity periods) and psychometric indicators based on the UTAUT model (Performance Expectancy, Effort Expectancy, Facilitating Conditions, Perceived Risk). All behavioral parameters are quantified using a 5-point Likert scale.A variable measuring inactivity during three months is introduced and used to develop an indicator for supervised learning tasks, including binary classification of churners and non-churners.The provided dataset can be used for machine learning purposes, particularly to model user behavior and to predict churn.The dataset serves as a region-specific basis for research in fintech environments and the promotion of digital financial inclusion."