The mentor gap

I have been mentored, and I have mentored, and I believe in it enough to have written a book that is partly an attempt to do it at scale. So I want to be careful not to oversell it, because the numbers in circulation on this subject are some of the worst I have checked.

What is real

Those are large, consistent differences on how work feels. On the objective outcomes, the picture is more sober. A meta-analysis covering 112 studies found workplace mentoring associated with career recognition and success at a correlation of about .05, and performance at about .06. Those are small. The effects on attitudes and satisfaction, at .14 and .16, were meaningfully larger.

An earlier meta-analysis found promotions correlating at .31, but that estimate rested on only three studies, which is thin enough that I would not build a plan on it.

The same survey found mentored workers likelier to feel fairly paid, 79% against 69%, and far likelier to believe their employer offered a path forward, 71% against 47%. Whatever a mentor does to promotion rates, it changes what people believe is available to them.

What is not real

You will encounter a claim that mentoring increases promotion and retention of underrepresented employees by 15 to 38%, usually attributed to Cornell. It traces to a graduate course assignment, not a study, and it restated a figure about representation in management as though it described promotion rates. Every article repeating it inherited the error.

You will also encounter dramatic corporate case studies, most famously one claiming mentees were promoted five times more often. That comes from an unreplicated 2006 vendor analysis with no control for the obvious problem: people who seek out mentors differ from people who do not, in exactly the ways that predict promotion anyway.

The defensible diversity finding is different and better sourced. Analysis of more than 800 US firms across three decades found mentoring programmes raising representation of several underrepresented groups in management by 9% to 24%.

The finding that should change what you do

One meta-analysis separated informal mentoring from formal company programmes. Informal relationships produced a significant effect. Formal assigned programmes produced an effect that was not statistically distinguishable from zero.

Read that again if you work in a career services office. The benefit appears to attach to relationships people build, not to relationships an organisation assigns.

What to do with this

1. Stop waiting to be assigned one. The evidence says the assigned kind is the kind that does not work.
2. Do not ask anyone to be your mentor. It is a large request that obliges a stranger to define a relationship. Ask a specific question instead, and see whether a second conversation happens.
3. Accept that most will not convert, and that this is fine. Mentorship is the surviving remnant of many ordinary professional conversations.
4. Be useful in the other direction. The people who get mentored well are almost always the people who are worth spending time on.

This is the practical heart of the exposure gap. Access to informal relationships is distributed by accident of birth, and informal relationships are the ones that measurably help.