What do I do when my PhD supervisor is not supportive?

The four failure modes need different responses, so it is worth being precise about which one you have.

Before you pick a response, write down what is missing in observable terms. Jabre and colleagues define supervisory fit through concrete dimensions: how much supervision you want, career guidance, respect for nonacademic goals, working style, and the kinds of students who do well in that group (Jabre et al., 2021, p. 2). That is guidance rather than a controlled study, so treat it as a checklist and not as a measurement. Name the specific pattern you are living with, such as missed feedback, unclear expectations, or dismissal of your career plans, then compare it against the dimensions on the list (Jabre et al., 2021, p. 4).

The absent supervisor does not reply and cancels meetings. The response is structure and persistence: fixed recurring slots, short written updates whether or not they reply, and a second reader for feedback. Absent supervisors are often overloaded rather than uninterested, and a specific request with a deadline usually gets answered when a general one does not.

The unclear supervisor gives feedback you cannot act on. "This needs more depth" tells you nothing. The response is to convert vagueness into specifics in the room: which section, what would depth look like here, can you name a paper that does it well. Ask until you have something you could hand to someone else as an instruction.

The overloaded supervisor is supportive and has no time. The response is to make your work cheap to engage with: short documents, specific questions, and a clear ask at the top of every email.

The hostile supervisor is a different category and does not respond to any of this. If you are being belittled, threatened, or made to feel unsafe, that is a matter for your graduate school or research integrity office rather than a communication problem to solve.

Across all four, the protective move is the same: do not let one person be your only source of feedback. A second reader, a peer group, or a co-supervisor changes the dynamic and reduces the risk to your project.

Your thesis committee is the ready-made version of that second source. Gu and Bourne recommend using committee members as secondary mentors for research guidance, professional development, and disagreements with the primary mentor, and they advise keeping those members informed of your progress rather than contacting them only when something breaks (Gu and Bourne, 2007, p. 2046). They also note that committee practice varies, because some institutions form the committee early and others only near the end, so check which applies to you. If you have no committee, Gaillard and colleagues argue that informal contacts count, and they add a qualification the other sources leave out: being open about difficulty is not equally safe for everyone, and researchers often fear a supervisor's disapproval, including over plans for careers outside academia (Gaillard et al., 2022, p. 1).

How do I run a supervision meeting that goes somewhere?

Send an agenda two days ahead with specific questions and any material to read. Ask decision questions rather than open ones. Then email a summary of what was agreed within a day. That record is what turns vague encouragement into something you can act on and refer back to.

Decision questions do most of the work. "Should chapter three come before or after the methods discussion, and why" gets an answer. "What do you think of the draft" gets a general reaction you cannot use.

The written summary is also your protection. Six months on, a disagreement about what was advised is resolved by an email nobody disputed at the time.

When should I change supervisor?

When the relationship is harming the work and structure has not fixed it, when your supervisor lacks expertise the project needs, or in any case involving bullying or harassment. Changing is more common than people think and is normally handled by a graduate school rather than negotiated directly.

Talk to the graduate school or a director of postgraduate studies before doing anything else. They have handled this before, they know the procedure, and an early conversation is usually confidential.

Treat your health as a reason to act, not as a test of your resilience. Jabre and colleagues report that graduate students have disproportionately higher rates of depression and anxiety than the general population, and they recommend judging the whole doctoral environment, including work-life balance, financial stability, family and childcare support, and how the institution treats international and underrepresented students (Jabre et al., 2021, p. 5). Gu and Bourne say it more bluntly, that graduate school is demanding mentally and physically, that health comes first, and that a good mentor recognizes the student comes first as well (Gu and Bourne, 2007, p. 2045). Note that none of these sources names an ombudsperson or a formal complaint route, so check what your own institution actually offers instead of assuming a standard procedure exists (Gu and Bourne, 2007, p. 2046).

Adding a co-supervisor is often the lower-friction option and solves most expertise gaps. It is easier to arrange, less disruptive, and does not require anyone to declare that a relationship failed.