What do we expect from a good PhD supervisor?
Scientific guidance, certainly. Help with publications. Advice when experiments fail. Introductions to international colleagues. Support in developing an independent research identity. But the list rarely stops there.
Today, we also expect supervisors to explain funding systems, advise on international mobility, help students navigate academic recruitment, discuss careers beyond academia, build professional networks and, increasingly, support doctoral candidates through the personal pressures of an uncertain job market.
At some point, we should ask a simple question: are we expecting one person to do too much?
The supervisor at the centre of everything
The traditional relationship between a doctoral candidate and an experienced supervisor remains one of the great strengths of doctoral education. It is difficult to imagine rigorous research training without it. A good supervisor does far more than oversee a project. They teach methods, help formulate scientific questions, refine manuscripts, introduce students to a disciplinary community and gradually encourage them to think and act as independent researchers.
The problem is not that supervisors are failing to do their job. Quite the opposite. Many invest enormous amounts of time, experience and personal commitment in supporting their students.
The problem is that the job itself has quietly expanded. Research careers have become increasingly international, competitive and complex. Doctoral candidates must learn not only how to conduct excellent research, but also how publication strategies work, when to start thinking about grants, how to build professional networks, whether and when to move abroad, and how recruitment systems differ between countries.
And who explains all this?
In many doctoral programmes, the answer is still: the supervisor.
We have gradually transformed one person into scientific adviser, career mentor, grant consultant, network builder, project manager and, sometimes, first-line emotional support.
None of these needs is unreasonable. Expecting one individual to meet all of them may be.
The lottery of the “right” supervisor
There is also a question of fairness.
No matter how brilliant or dedicated a supervisor may be, the advice they can provide is inevitably shaped by their own career, discipline, experiences and professional network.
One doctoral candidate may happen to find a supervisor who understands international recruitment, actively introduces students to colleagues, discusses funding opportunities and openly explains the unwritten rules of academic life. Another, equally talented, may have an excellent scientific supervisor who provides little career guidance.
The difference may have nothing to do with merit.
Yet it can profoundly influence what opportunities each student sees and which decisions they make.
Good career preparation should not depend on individual heroism, or on the luck of meeting the “right” supervisor.
The limitation becomes even clearer when we look beyond academia. Only a minority of doctoral researchers will ultimately obtain permanent academic positions, yet we often expect academic supervisors to advise students about careers they may never have experienced themselves.
A professor can be an outstanding mentor and still know relatively little about careers in industry, government, entrepreneurship, science policy or other research-intensive environments.
That is not a failure of supervision. It is a limitation built into the model.
Sharing responsibility
The solution is not to weaken the supervisor–student relationship. It is to support it.
Doctoral schools could provide a small number of structured career conversations shared by all students: how research careers develop, how scientific independence is built, how publication and funding systems work, why professional networks and international mobility matter, how recruitment operates and what opportunities exist beyond academia.
These conversations do not need to become another heavy layer of compulsory training. Even a few well-designed sessions, introduced early in the PhD and revisited as students progress, could provide a common foundation.
Individual supervisors would then be able to do what only they can truly do: interpret that general knowledge within a specific discipline, discuss individual choices and provide the scientific judgement that develops through years of experience.
Institutional career education and individual mentoring are not competitors. The first can make the second considerably more effective.
From academic careers to research careers
There is a broader reason why this matters.
For too long, doctoral career discussions have implicitly treated academia as Plan A and everything else as Plan B. But for most PhD graduates, a permanent academic career will not be the eventual destination.
That does not make the doctorate a failed investment, nor does it make careers outside universities second choices.
The skills developed through research, analysing complex problems, writing clearly, managing long-term projects, working with uncertainty, learning new fields quickly and collaborating across disciplines and countries, are valuable in many professional environments.
The purpose of career preparation should therefore not be to persuade doctoral students either to remain in academia or to leave it. It should be to help them understand the landscape early enough to make informed choices about where their abilities, interests and ambitions can best develop.
Supervisors will remain indispensable to doctoral education. But they should not have to compensate for everything doctoral education does not formally provide.
Perhaps the best way to strengthen PhD supervision is to stop asking supervisors to do the entire job alone.