The greatest footballer of any generation is not allowed to coach a professional team. Not without a licence. European football requires years of staged coaching education, the UEFA licence ladder, before anyone, however gifted a player, may take charge of other players’ development, because the sport learned an expensive lesson early: playing and coaching are different crafts. The instincts that made someone brilliant on the pitch transfer far less than anyone assumes, and a dressing room run on “just do what I did” produces confusion in the many players who are not built like the coach. So sport separated the two careers, built a curriculum for the second one, and made it mandatory. The result is unremarkable only because it works: nobody in professional sport finds it strange that developing performers is a licensed skill.
Academia runs the experiment sport abandoned. It identifies its best players, promotes them directly into coaching, and issues no licence, no curriculum, and no apprenticeship in the craft, only a lab full of developing performers and the silent assumption that research excellence includes them. I have written about how that experiment goes, and the literature says the same: supervision style is among the strongest predictors of doctoral mental health [1], students’ perception of their supervision tracks their burnout directly [2], and almost every difficult supervisor a student describes is not malicious but unlicensed. This article is the missing first course: what the supervision craft actually consists of, in five learnable moves.
Table of Contents
- Daniel supervises the way he was supervised
- The craft, not the folklore: five moves
- “They do not really know how to do it”: the licence, requested from below
- What Daniel’s student notices first
- A word for the coordinator who owns the licence problem
Daniel supervises the way he was supervised
Daniel, eight months into running his group, is supervising his first doctoral student, and under pressure he has defaulted to the only curriculum he ever received: his own supervisor’s behaviour. He is unavailable for stretches, then suddenly intense; his feedback arrives as “needs work” because that is what feedback looked like when he received it; he assigns stretch tasks without checking whether the student can yet swim at that depth, because that is how he learned, and he survived. This is the quiet mechanism by which supervision styles propagate through academia like folklore: each generation coaches from a sample size of one, their own experience, edited by survivorship. Daniel is not a bad supervisor. He is an unlicensed one, running the only playbook he was ever shown, and his student, who is not built like Daniel, is drowning politely.
The craft, not the folklore: five moves
The good news, and it is real, is that the supervision craft is smaller than it looks from inside the fog. Five moves cover most of it, and every one is learnable in weeks, not years.
Move one: diagnose before you delegate. The core coaching skill is matching your style to the person’s current development stage on the specific task, directive for the new, coaching for the developing, delegating only to the truly ready, and mismatches in both directions do damage: micromanaging the competent, abandoning the beginner. This diagnostic is a full framework of its own, and it has a home of its own: situational leadership in academia, which is the closest thing to the licence’s first module that we can offer you in an afternoon. Read it before anything else here.
Move two: install the interface, from above. In my article on the PhD supervisor relationship, students are taught to build the supervision interface from below: agendas sent before meetings, written summaries after, feedback requests specific enough to answer, an agreed default when silence blocks a decision. Every one of those works better installed from your side, because you have the authority to make it culture rather than initiative. One page of structure per student per week converts your meetings from status rambles into decision sessions, and it is the cheapest hour of leverage in your calendar.
Move three: feedback as specification, never verdict. The word feedback has been mugged on its way through academia, and this move takes it back. In engineering, feedback is not an opinion. It is output fed back into a running system so the system can correct course: the thermostat’s reading, arriving in time to adjust the boiler. Steering information, by definition. What students get instead is a verdict after the fact. “Needs work” steers nothing; to a developing researcher it reads as “not yet worthy”, and it buys you a month of invisible, anxious polishing you never asked for. The craft version restores the literal meaning, and costs one sentence more: what specifically, by when, and to what standard of done, delivered inside an agreed window, while the work can still change course. Feed it back. That is the entire word. Decision-level feedback is the single behaviour that most changes how students describe their supervision, and you can adopt it this afternoon.
Move four: run the wellbeing radar, and know its limits. One in two doctoral students experiences psychological distress, so a supervisor who cannot recognise the early pattern, exhaustion that rest stops repairing, withdrawal from the group, a competent student suddenly re-checking everything, is coaching with eyes closed; I walk through those signals one by one in my post on PhD burnout. But recognising is the whole of your clinical job. You are a coach, not a therapist: the craft move is naming what you observe, kindly and privately, and routing to real support, the counselling service, a doctor, the graduate school, never diagnosing, never treating, never making your concern a burden the student must now manage. If this area touches things that feel personally close, for you or a student, involving proper support early matters more than anything else in this article.
Move five: say aloud that you are learning the craft. The highest-leverage sentence available to a new supervisor is some version of: I was never trained for this part of the job, I am learning it deliberately, and I want your feedback on how it is working. It costs one moment of ego and buys the thing the whole relationship runs on, which is permission for frankness in both directions. Licensed coaches are observed and assessed throughout their careers; the unlicensed can at least volunteer for it.
“They do not really know how to do it”: the licence, requested from below
If you doubt that any of this is visible from the other side of the desk, here is a participant in one of our trainings, describing their own supervisors and then doing something remarkable about it:
“The problem of many researchers is that they are concentrated only on their field of study. Sometimes, I see that supervisors want to support their students, but they do not really know how to do it. Therefore, I am eager to learn more about the mechanisms of work-related psychological issues to help my future students to avoid burnout.”
Read the three sentences in order, because together they are this entire article in a participant’s handwriting. First, the diagnosis, and notice its generosity: not “my supervisors are bad people” but they want to support their students and do not know how, which is the unlicensed-coach problem named cleanly, goodwill without craft. Second, the mechanism: concentrated only on their field of study, which is just what a system produces when it promotes players and never teaches coaching. And third, the part I find quietly moving: this researcher, before ever holding the role, is requesting the licence, deliberately acquiring the craft for students who do not yet exist. The folklore chain breaks right here, one person deciding to be trained instead of merely descended from. Every institution has researchers like this one. The question the close of this piece asks is what their institution is doing to meet them.
What Daniel’s student notices first
Daniel, to his credit, did the thing the fifth move describes: he told his student he was learning the job. Then he installed the interface, one agenda, one summary, a two-week feedback window, and read the situational framework twice. The change his student noticed first was not warmth, and this is worth every new supervisor hearing: it was predictability. Feedback that arrives inside a known window, expectations written where both can see them, a supervisor whose availability has a shape, these did more for the student’s composure than any amount of encouragement, because a developing researcher can adapt to almost any coach whose behaviour is legible. Folklore supervision is not usually cruel. It is arbitrary, and arbitrariness, not strictness, is what corrodes people. Six months on, Daniel’s supervision is still imperfect and no longer inherited. His student, meanwhile, has started sending agendas to their own undergraduate project student, unprompted, which is how the better folklore starts.
A word for the coordinator who owns the licence problem
And if you run researcher development, the sport comparison hands you the institutional question in one line: your organisation would never let an unlicensed coach near a youth academy, and it hands every new group leader a lab full of doctoral students on the strength of their playing career. The supervision literature tells you what that costs, in your students’ mental health and your completion rates, and the engineering answer is the same one sport found: make the craft trainable, early, and normal, ideally in the window where future supervisors are already requesting it, like the participant above, late doctorate to early group leadership, before the folklore sets. Supervisor development is the highest-leverage line in a researcher-development budget, because every trained supervisor upgrades every student they will ever have. The situational toolkit, the interface, feedback craft and the wellbeing radar are what we build with supervisors and future supervisors across our trainings, with over 1,500 individuals trained at more than 30 institutions; book an intro call and ask us the licensing question directly. Sport stopped assuming its best players could coach half a century ago. Academia’s players deserve the same seriousness, and so, rather more urgently, do their students.
FAQ
How do I supervise a PhD student for the first time?
Start by dropping the inherited playbook: supervising the way you were supervised is coaching from a sample size of one. Five moves cover the craft: diagnose each student’s development stage before choosing your style; install a structured interface (agendas, written summaries, feedback windows); give decision-level feedback, never verdicts like “needs work”; learn the early wellbeing signals and route to proper support; and say aloud that you are learning the job.
Why is PhD supervision so often bad?
Rarely from malice, almost always from the licence gap: academia promotes its best researchers into a coaching role with no training, so each generation supervises from folklore, their own supervision, edited by survivorship. Goodwill without craft produces just what students report: vague feedback, arbitrary availability, mismatched autonomy. The fix is training, and a 2017 study of more than 3,600 PhD students linked supervision quality directly to doctoral mental health.
What is the most important supervision skill?
Matching your style to the student’s development stage on the specific task: directive where they are new, coaching where they are growing, delegating only where they are ready. Mismatches do damage in both directions, micromanaging the competent and abandoning the beginner. The framework is called situational leadership, and its academic translation is the first module of the craft.
What should a supervisor do if a PhD student is struggling emotionally?
Recognise and route, never diagnose or treat. Name what you observe privately and kindly, then connect the student to real support: the counselling service, a doctor, the graduate school. A supervisor’s clinical job ends at noticing early and making the path to help easy; taking on the therapist’s role helps nobody and risks real harm. Early, low-drama routing is the craft move.
References
1. Levecque, K., Anseel, F., De Beuckelaer, A., Van der Heyden, J., & Gisle, L. (2017). Work organization and mental health problems in PhD students. Research Policy, 46(4), 868–879.
2. Cornér, S., Löfström, E., & Pyhältö, K. (2017). The relationships between doctoral students’ perceptions of supervision and burnout. International Journal of Doctoral Studies, 12, 91–106.
Last Updated on August 12, 2026 by Dr Nadine Sinclair
Dr Nadine Sinclair
Dr. Nadine Sinclair is a molecular biologist, strategy consultant and co-founder of Mind Matters. A scientist by training and at heart, she conducted her doctoral research at the Max Planck Institute for Biophysical Chemistry. She has 30,000+ hours of hands-on project management experience, built across 18 years as a strategy consultant at McKinsey and as an independent consultant, working with research institutions, foundations, pharmaceutical and biotech companies and governments. She is the co-creator of the Personal Resilience Indicator and the author of On Track. Since 2018, over 1,500 researchers from more than 30 institutions have trained with Mind Matters.