Robo-advisor acceptance: Do gender and generation matter?

Robo-advice technology refers to services offered by a virtual financial advisor based on artificial intelligence. Research on the application of robo-advice technology already highlights the potential benefit in terms of financial inclusion. We analyze the process for adopting robo-advice through the technology acceptance model (TAM), focusing on a highly educated sample and exploring generational and gender differences. We find no significant gender difference in the causality links with adoption, although some structural differences still arise between male and female groups. Further, we find evidence that generational cohorts affect the path to future adoption of robo-advice technology. Indeed, the ease of use is the factor which triggers the adoption by Generation Z and Generation Y, whereas the perceived usefulness of robo-advice technology is the key factor driving Generation X+, who need to understand the ultimate purpose of a robo-advice technology tool before adopting it. Overall, the above findings may reflect that, while gender differences are wiped out in a highly educated population, generation effects still matter in the adoption of a robo-advice technology tool.

1. While the gender comparison did not produce notable differences, the paper does describe some minor gender gaps that may still require attention. I suggest that the authors note finding minor gender differences and stating their nature in the abstract, as just saying "we do not find significant differences between male and female" may lead readers to think that no differences were found.
Indeed, we find some differences, but they are not statistically significant. We agree that our sentence in the abstract can be confusing and we have rephrased it to avoid any misunderstanding. Thank you for pointing this out.
2. Overall, it is interesting -refreshing yet surprising -that no significant gender gaps were found, as much of research on new technology adoption in the recent years show that men are more accepting of new technologies than women in various domains. Is there something that's different about the nature of robo-advisors that may contribute to the lack of differences? Could it be a result of the sample composition, e.g., did males and females in the sample share comparable characteristics, or were there characteristics of the sample that may have contributed to the muted gender effects?
You are right, the results are a consequence of the special characteristics of the sample (the high education of the respondents), that, in our opinion, contribute to levelling the gender gap. We have better highlighted this aspect of the research.
3. The very last paragraph stating limitations of the study needs to be elaborated much further. The sample composition poses additional challenges and limitations than what the authors have mentioned, as there are additional characteristics that make the sample far from representative. A deeper discussion of the sample limitations, and how these may have contributed to the findings need to be discussed. Additionally, it would be helpful to see examples and suggestions around the "behavioral studies" proposed to improve the objectivity and validity of the research.

We have elaborated more on the limitations of the study in the concluding section.
Regarding the possibility to carry on an analysis without the limit of self-reported data, a possible approach is to perform virtual experiments on (virtual) portfolio management where respondents are guided through specific choices, also involving robo-advisory possible applications (behavioral finance approach). Alternatively, one can possibly request real data from specific financial platforms supporting the use of robo-advice, such as in Bianchi-Brière (2021), to analyze how investors manage their portfolio in reality. Of course, these are sensible data and not easy to obtain. Anyway, since we realized that such a comment did not add much to the discussion of our results, we deleted the reference to behavioral studies. 4. While the paper is written in standard language and does not have major grammatical errors, use of non-academic language and subjectively toned sentences are noticeable throughout the discussions which seem unfit for a research manuscript.
The revised paper has undergone a full proofreading process with a professional. We hope that any inappropriate language has been removed from the text.