Up until now, research on the influence of mentorship on the career perspectives in young researchers is scarce. It seems that, in academia, there is a consensus that mentorship is beneficial in general, but the extent to which the positive effects of mentorship hold in the long term is not clear. According to Eby et al. 2008, mentorship has moderate benefits in terms of health-related, motivational, and career outcomes. However, this much depends on the type of mentorship. According to the Nature Neuroscience editorial from 2007, mentorship is highly recommended in an academic environment, and can be productive even when there is a long geographical distance between the mentor and the mentoree.
However, I have a feeling that the topic has not been covered as in depth as it deserves. Hereby, I would like to briefly discuss the following three aspects of the issue:
 A cloud rather than a ladder
First of all, the same as there are multiple types of supervisors, there are also multiple types of mentors. This means that, the same as in research, in mentorship there can also be a mismatch between two people. As people have a tendency to generalize, some successful mentors interpret their own coping strategies as the preferred way of living, even though there is not just one way of becoming successful or happy. Moreover, there was some research on the influence of having a single mentor, or a role model, on achieving personal goals. According to this research, mimmicking your idol’s behaviours can, in a long run, pose obstacles to the development of your own identity, and decrease the chances for achieving a long time success. What is the solution then? I guess, the solution would be building a decentralised network, where one mentee has access to multiple mentors. This solution sounds very obvious, however I do not see this solution implemented. A ‘mentor’ is still perceived as a form of a general practitioner or a couch, who is scheduled for the meetings on a regular basis. While in fact, a mentorship system should be more of a cloud, where mentees would have plenty of possibilities - any time any place - to reach out to more experienced researchers for help.
 Mentor versus mentee
Secondly, a stiff division into a mentor and a mentee can be highly counterproductive. It was demonstrated on multiple occasions that mutual mentoring in closed groups works best. In this system, the participants mentor their peers, therefore, everyone becomes a mentor and mentee at the same time. Being a mentor to fellow students has a range of positive benefits on the level of self-reflection, motivation and a sense of self-value. Mentoring others causes shifting attention from our own problems to other people’s issues, and helps in getting a distance from your own situation. When I experience a bad day at work, what best works for me is checking what my students are doing at the moment, and trying to solve issues appearing at other projects I am participating in, rather than staying focused on solving my own issues. Furthermore, mentoring others helps you realise that everyone has their own issues, regardless of how much successful they seem at the first glance. And that there is nothing like a comfort zone in academia. Therefore, mentoring and supervising others should be a natural part of the PhD training process, next to writing research papers, presenting at conferences and writing a thesis.
 A sense of success
Lastly, it is still highly uncommon to treat your mentorship experience as a personal success. I have never seen a mention about a successful mentorship in any CV. Some general change in thinking about mentorship is necessary, as the ability to mentor others should be treated as a badge of honour rather than as a little extracurricular activity, negligible in a CV.