Most mentoring relationships do not end. They just stop being scheduled. The standing coffee moves once, then twice, then becomes a text thread, then becomes nothing. Nobody declares failure, because nothing visible failed. Both people were sincere. Both showed up for months. And yet if you asked either of them what changed as a result, you would get a vague answer about perspective.
I have been on both sides of that. I have been the mentee who took a year of someone's generosity and converted almost none of it into a decision, and I have been the mentor who mistook being pleasant company for being useful. The reason it happens so often is that we carry a romantic idea of what mentorship is — the seasoned hand transferring wisdom to the eager successor — and that idea does not describe what actually moves between two people. Effective mentorship in business is narrower and less flattering than the mythology. It is also far more useful once you accept the narrowness.
What actually moves between two people
The first transferable thing is pattern recognition. Not knowledge, not judgement, not experience in the abstract — pattern recognition. The mentor has seen a particular shape of problem enough times to recognise it early, and the mentee has not. When somebody describes a founder dispute, a stalling sales cycle, a leadership hire who interviews brilliantly and lands badly, the useful thing I can say is rarely "here is what to do." It is "I have seen this shape before, and here is where it usually goes next, and here is the thing that looked irrelevant at the time and turned out to be the whole problem."
That is a genuine transfer, because the mentee cannot manufacture it. Pattern libraries are built by living through outcomes, and living through outcomes takes years. A mentor can compress the timeline — not by handing over conclusions, but by lending the shape. The mentee still has to test whether their situation matches. Often it does not, and the honest mentor says so quickly rather than forcing the fit.
Permission is the underrated half
The second thing that transfers is permission, and it is doing more work in these relationships than anyone admits. A striking share of the people who bring me a problem already know the answer. They know they need to exit the customer segment, or replace the person they hired, or stop chasing a market that is politely ignoring them. What they lack is not analysis. It is the standing to act on their own conclusion without feeling reckless.
When someone senior says "yes, that is what I would do too," a decision that felt like a gamble becomes a defensible judgement. That is not weakness on the mentee's part. Most people in their first decade of responsibility are operating without an internal reference point for how much conviction is normal, and they read their own uncertainty as evidence that they are wrong. A mentor's role is often to say: this level of doubt is standard, act anyway.
The risk is obvious. Permission is easy to give and easy to give badly. I try to withhold it until I understand the decision well enough to be genuinely willing to be wrong alongside them. If I am not prepared to hear how it turned out, I have no business ratifying it.
The parts that never transfer
Almost everything else people hope to pass along stays put. Tactics do not travel, because tactics are bound to a moment — a market condition, a capital environment, a particular team's tolerance. What worked when I was running consulting teams would be an odd thing to hand to someone building today, and repeating it as instruction rather than as an anecdote is how mentors become tedious.
Temperament does not transfer either. Some people can hold an unpopular position for eighteen months without flinching; others cannot, and no amount of coaching converts one into the other. The useful move is to help someone design around their temperament rather than apologise for it. Risk tolerance is the same. My comfort with a particular kind of exposure was earned through a specific set of experiences and is not advice.
Network transfers only partially. I can make an introduction, and that is real value. But the relationship that made the introduction possible is not something I can hand over. The mentee inherits a warm door, not the standing behind it, and confusing those two is how a good introduction gets burned.
And your path does not transfer at all. This is the failure I see most in structured leadership development programmes, where the implicit model is that the senior person's career is a template. It is not. It is one sample, heavily shaped by conditions that no longer exist.
How these relationships fail without anyone noticing
The quiet failure has a recognisable rhythm. The mentee arrives with a status update rather than a decision. The mentor listens, finds it interesting, and responds with a story from their own career that is adjacent but not applicable. Both parties leave feeling that something worthwhile happened. Nothing did.
The diagnostic I use is blunt: if I cannot name the decision the other person is facing, the meeting has not started yet. Not the situation, not the context — the decision. "Do I keep this person or replace them by the end of the quarter" is a decision. "Things are hard with the team" is a weather report. Getting from one to the other is the mentor's job, not the mentee's, because the mentee usually does not know that framing the question is the work.
The second failure mode is cadence without content. A monthly meeting that exists because it exists will drift toward pleasantness, and pleasantness is the enemy here. I would rather have four sharp conversations a year, each triggered by something real, than twelve comfortable ones. That principle holds in governance too — I have argued elsewhere that the substance of board work happens between the meetings, not in them, and the same is true of a mentor mentee relationship.
Working at the right altitude
The most common mistake I make as a mentor is answering at the wrong altitude. Someone brings me a question about how to structure a difficult conversation, and I respond with a view on their five-year positioning. It feels generous. It is actually a way of avoiding the specific question, which is harder and requires me to know details I have not bothered to learn.
The reverse happens too. Someone raises a strategic question and I retreat into tactics, because tactics are where I feel competent. Learning to notice which altitude the question was asked at — and to answer there before climbing — is most of what improved my own practice.
This is also why I take fewer people than I am asked to. The founding work I did with GLOBE, and the time I spend on a board at the FIU Honors College, taught me that attention is the scarce input, and diluted attention produces exactly the pleasant, useless relationship described above. I do the same thing with mentoring that I do before accepting a board seat: I ask to observe first. A single conversation about a live problem tells you whether you have anything to offer this particular person. If you do not, saying so early is a kindness. The logic is identical to deciding which customers not to serve — capacity spent on a poor fit is capacity stolen from a good one.
What should survive
There is an asymmetry in these relationships that should make every mentor careful. The mentee remembers sentences the mentor has entirely forgotten saying. Offhand remarks land as verdicts. I have had people repeat something back to me years later that I said in passing and would not have defended if pressed. Anyone thinking about how mentorship builds the next generation of business leaders should sit with that for a moment. Casual authority is still authority.
So the test of whether it worked is not gratitude, and it is not whether the person's career went well — plenty of things determine that, and you are a minor one. The test is whether they left with your questions rather than your answers. If, two years on, they catch a problem earlier than they would have, and back their own read on it without needing anyone to sign off, the transfer happened. That is the whole of it. We do not pass on wisdom. We pass on a few shapes worth recognising and the nerve to act on what we see.