Drivers of robo-advisors adoption in financial services: a meta-analytical study

Purpose This study aims to synthesize existing empirical findings to identify the primary drivers of robo-advisor (RA) adoption in financial services. It addresses inconsistencies in the literature regarding key predictors – such as trust, perceived risk and usefulness – and investigates how national, economic and cultural factors moderate these relationships. By examining 41 studies across more than 20 countries, the research provides a comprehensive, evidence-based understanding of RA adoption behavior in diverse financial and technological environments. Design/methodology/approach Using a meta-analytical method, the study aggregates data from 41 empirical studies comprising 17,558 respondents. A systematic search and coding protocol were followed, and effect sizes were computed and analyzed using random-effects models. Moderator analyses were performed using hierarchical meta-regression to assess the influence of gender, the Human Development Index, Country Innovation Index and Hofstede’s cultural dimensions on adoption drivers. Statistical tests for heterogeneity, publication bias and model stability were also conducted to ensure robustness and reliability. Findings The meta-analysis revealed that attitude, trust, perceived usefulness and ease of use are the strongest predictors of RA adoption. Contrary to prior assumptions, financial literacy and perceived risk showed no significant effects. Moderation analysis highlighted that women perceive greater usefulness of RAs and that national development and innovation levels enhance adoption effects. Additionally, cultural values such as individualism, long-term orientation and indulgence positively influence the strength of perceived usefulness and adoption intention relationships, emphasizing the importance of context-specific adoption strategies. Originality/value This study is among the first to meta-analytically consolidate global evidence on RA adoption, incorporating cross-national and cultural moderators. It advances theoretical models like TAM and UTAUT by embedding boundary conditions from cultural and economic contexts. Practically, the findings offer actionable insights for fintech developers and financial institutions to tailor robo-advisory platforms based on demographic and cultural characteristics. The study also challenges traditional gender-based technology adoption assumptions, suggesting greater algorithmic trust among women in financial technology applications.