Most managers who mentor are making it up as they go.
Here is what the research shows.
Five studies, one uncomfortable picture, and why "willing senior managers" is not a development plan.
The line on the home page is meant to be slightly provocative. It is also, on the evidence, accurate. In most organizations, the people doing the mentoring are senior managers who were nominated because they had time, seniority, or a willing temperament; rarely because anyone had assessed whether they could mentor. They turn up to the first session and improvise. Sometimes that works. Often, it does not, and the reasons are not what people assume.
The temptation, when a mentoring relationship goes wrong, is to look for a single failing. A bad match. A busy mentor. A mentee who would not engage. The research tells a more uncomfortable story: mentoring fails for several reasons at once, and no single study can capture all of them. What follows is a survey of five works that, taken together, explain why the home page claim is not hyperbole. Each study isolates a different facet of the problem. Each is referenced in the book. Together, they make the case that "throwing willing senior managers into the battle" is not a strategy; it is a hope.
1. Bad mentoring is more common than the brochure suggests
The most-cited single finding in the negative-mentoring literature comes from Eby, McManus, Simon and Russell (2000), whose taxonomy of negative experiences remains the reference point twenty-five years on. In a sample of working adults who had been mentored, more than half reported at least one substantive negative experience with their mentor. The experiences ranged from the relatively mild (a mismatch in working styles, a mentor whose expertise had become outdated) to the seriously damaging (manipulation, credit-taking, sabotage).
The point is not that most mentors are malign. Most are not. The point is that the base rate of mentees who walk away with at least one bad experience is around 50 percent, which is not the figure most organizations have in mind when they set up a program. And yet the assumption persists that having a mentor is, by definition, better than not having one. The data does not support that assumption.
2. Bad mentoring leaves people worse off than no mentoring at all
Hu, Kwan, Zhang and Li (2024) followed up the negative-experience literature with a longitudinal study of mentees who had reported negative experiences. Their finding was uncomfortable: those mentees showed higher turnover intention, lower job satisfaction, and greater psychological withdrawal from work than employees with no mentor. In other words, a poorly run mentoring relationship is not a neutral event from which the mentee bounces back. It is an active drag on the things mentoring programs are designed to improve.
A companion paper from the same research group (Hu, Li and Kwan, 2022) extended the analysis to mentors themselves, finding that mentors who experienced difficulty in the relationship suffered measurable reductions in creativity and increases in emotional exhaustion. Bad mentoring damages both parties. This is the asymmetry that the "willing senior manager" model misses entirely: the cost of getting it wrong falls on both the mentor and the mentee, and a depleted mentor is less useful to anyone else they are trying to help.
3. The dysfunction is structural, not personal
Scandura (1998) was one of the earliest researchers to argue that mentoring failure is not best understood as the work of bad apples. Her analysis of dysfunctional mentoring relationships identified five recurring patterns: negative relations, sabotage, difficulty, spoiling, and submissiveness. Crucially, she showed that these patterns emerged from the structure of the relationship itself, the power asymmetry, the absence of formal accountability, and the lack of an exit mechanism, rather than from the personalities involved.
This matters because it changes what counts as a fix. If dysfunction is personal, the response is to pick better people. If dysfunction is structural, the response is to design the surrounding scaffolding more carefully: clear contracting, defined boundaries, regular check-ins, and a route through which either party can raise a concern without it ending the relationship. Twenty-seven years on, very few in-house programs do any of this systematically.
4. Mentor training is not optional, and most mentors do not get it
Treasure and colleagues (2022), writing in PLOS Computational Biology, set out ten principles for establishing mentoring programs that work in practice. Their starting point is uncompromising: a program that pairs people, provides no training, and offers no oversight is not a mentoring program. It is a set of meetings with hopeful labels. They identify trained mentors, defined guidelines, regular evaluation, and an accessible reporting mechanism as the minimum infrastructure for a sustainable program; not as bureaucratic additions, but as the conditions under which good intentions can survive contact with reality.
The gap between this and standard practice is wide. Most managers who are asked to mentor have had no training in active listening, as it differs from giving advice, no training in the contracting conversation that opens the relationship, no training in how to give feedback that is heard rather than absorbed defensively, and no training in how to end a relationship that is not working. These are skills. They can be taught. Hardly anyone teaches them.
5. When mentoring is done well, it works. When it is not, it does not.
The counterweight to all of this comes from the meta-analytic literature. Eby, Allen, Evans, Ng and DuBois (2008) pooled findings across more than a hundred studies and confirmed that mentoring, when it goes well, produces moderate but real effects on career outcomes, job satisfaction, and retention. A more recent meta-analysis from the same research line (Eby, Allen, Evans, Ng and DuBois, 2021) confirmed that the effects hold across sectors and decades. Ghosh and Reio (2013), in a separate meta-analysis, showed that the benefits extend to mentors themselves, who report higher career success, stronger job attitudes, and better organizational commitment when their mentoring relationships are working.
The phrase to underline in all of these studies is "when it goes well". The effects are real, but they are conditional. A mentoring relationship that has been thrown together, given no training, no contracting, and no oversight, has roughly the statistical profile of a coin toss: as likely to harm as to help. The averages reported in the meta-analyses are positive because some programs are well-designed. The averages mask a wide variance, and that’s the point.
What the five studies, taken together, show
There is no single failing here, and that is the lesson. Mentoring goes wrong in at least five distinguishable ways: mismatched pairs, distancing behavior, manipulation, gaps in mentor expertise, and general dysfunction in the relationship dynamic. Each of these has its own causes, warning signs, and remedies. No single intervention addresses more than one or two of them. This is what makes the "willing senior manager" model so inadequate: it assumes the only requirement is good intent, when the evidence shows that good intent is almost incidental to whether the relationship works.
A more constructive response starts from a different premise. Mentoring is a discipline. It has a body of skills that can be named, taught, and practiced. It has predictable failure modes that can be anticipated and designed around. It has measurable outcomes that can be tracked. None of this is exotic; it is the same standard we would apply to any other capability we expect senior managers to deploy. Yet for reasons that are partly historical and partly cultural, mentoring has been left in the realm of the gifted amateur, with predictable results.
The home page line, "most managers who mentor are making it up as they go", is not a complaint about managers. It is a description of an organizational pattern. The fix is not to find better mentors. It is to stop pretending that mentoring is something a sufficiently senior person can simply pick up.
Where the book takes this further
These five studies are the entry points, not the destination. From Manager to Mentor: Conversations You Cannot Rehearse draws on these and around forty other research sources to set out, chapter by chapter, what good mentoring involves in practice: the contracting conversation, the questions that open thinking, the feedback that lands, the difficult ending, and the dozen other moves that distinguish a useful mentor from a willing one. If the research surveyed here resonates with what you have seen in your own organization, the book is the next conversation.
References
Eby, L. T., Allen, T. D., Evans, S. C., Ng, T., & DuBois, D. L. (2008). Does mentoring matter? A multidisciplinary meta-analysis comparing mentored and non-mentored individuals. Journal of Vocational Behavior, 72(2), 254–267.
Eby, L. T., Allen, T. D., Evans, S. C., Ng, T., & DuBois, D. L. (2021). Does mentoring work? A meta-analytic review and research agenda. Journal of Vocational Behavior, 126, 103553.
Eby, L. T., McManus, S. E., Simon, S. A., & Russell, J. E. A. (2000). The protégé’s perspective regarding negative mentoring experiences: The development of a taxonomy. Journal of Vocational Behavior, 57(1), 1–21.
Ghosh, R., & Reio, T. G., Jr. (2013). Career benefits associated with mentoring for mentors: A meta-analysis. Journal of Vocational Behavior, 83(1), 106–116. https://doi.org/10.1016/j.jvb.2013.03.011
Hu, Z., Kwan, H. K., Zhang, Y., & Li, J. (2024). The effects of negative mentoring experiences on protégés’ turnover intention: The roles of harmonious work passion and moqi with the mentor. Journal of Managerial Psychology, 39(6), 716–731. https://doi.org/10.1108/JMP-01-2023-0017
Hu, Z., Li, J., & Kwan, H. K. (2022). The effects of negative mentoring experiences on mentor creativity: The roles of mentor ego depletion and traditionality. Human Resource Management, 61(1), 39–54.
Scandura, T. A. (1998). Dysfunctional mentoring relationships and outcomes. Journal of Management, 24(3), 449–467.
Treasure, A. M., Hall, S. M., Lesko, I., Moore, D., Sharan, M., van Zaanen, M., Yehudi, Y., & van der Walt, A. (2022). Ten simple rules for establishing a mentorship program. PLOS Computational Biology, 18(5), e1010015. https://doi.org/10.1371/journal.pcbi.1010015
