Avaliação de usabilidade em aplicativos bancários digitais: um estudo de caso do Caixa Tem

Abstract

This study analyzes usability and user experience in digital banking applications through the integrated application of Human–Computer Interaction methods, using the Caixa Tem application as a case study. The investigation is motivated by the widespread adoption of the platform in the context of public policies and by the mismatch identified between its social relevance and the levels of satisfaction perceived by users. The research adopts a qualitative approach, complemented by quantitative data, structured in sequential and integrated stages. Initially, a desk research was conducted to contextualize Caixa Tem within the digital banking application ecosystem. Subsequently, the MALTU methodology was applied to analyze spontaneous user comments published on the Google Play Store, enabling the identification of recurring interaction problems based on natural language. Afterwards, a heuristic evaluation was carried out by specialists, based on heuristics adapted for mobile applications, with the purpose of identifying usability violations and classifying them according to their severity level. Additionally, usability tests were conducted with real users, high school students who are beneficiaries of public policies, in which performance, error occurrence, and task comprehension were evaluated, both in lowf idelity wireframes and in medium-fidelity prototypes. The evaluation was complemented by the application of the System Usability Scale and the User Experience Questionnaire, allowing the analysis of subjective perceptions of the user experience. The results showed a predominance of negative manifestations related to efficiency, effectiveness, support, and security, as well as the identification of high-severity interface problems. Furthermore, the proposed redesign contributed significantly to reducing task execution time, decreasing errors, and substantially increasing usability and perceived experience scores. It is concluded that the integrated application of Human–Computer Interaction methods proved effective in identifying systemic usability problems and in supporting the proposition of design improvements in digital systems of high social relevance, fostering more efficient, intuitive, and reliable experiences.