Most candidates get somewhere between twelve and thirty supervision meetings a year. Used well, that is the mechanism by which a project stays on course. Used badly, it is a monthly status update in which you describe what you did, your supervisor nods, and nothing is decided.
The gap between those two is almost entirely procedural, and almost entirely within your control — which is fortunate, because you are the one with the most at stake.
Why meetings drift
Three failure modes account for most wasted supervision.
The meeting has no agenda, so it defaults to a progress report. Progress reports feel productive and decide nothing. You leave with encouragement rather than direction, and the actual blocker — which you did not raise because it was embarrassing — is still there next month.
Material arrives in the room. Handing over a chapter at the start of the meeting guarantees you get an impression rather than a considered read. Supervisors are not being dismissive; nobody can assess twelve pages of argument in four minutes.
Nobody writes anything down. Six weeks later you remember being told to narrow the sample; your supervisor remembers suggesting you consider it. Neither of you is lying. Both of you are reconstructing a conversation that was never recorded.
The agenda that fixes it
Send a short agenda 48 hours ahead, with any material attached. Four headings are enough:
- Since last time — three or four lines. What was agreed, what got done, what did not and why.
- Decisions I need — the core of the meeting. Frame each as an actual choice with options, not an open question. "Should I use purposive or theoretical sampling? Purposive fits my timeline; theoretical fits my grounded theory framing. I lean purposive because…" gets you a decision. "I've been thinking about sampling" gets you a discussion.
- Blockers — what is actually stopping you, stated plainly. This is the item candidates most often omit and most need.
- Next — what you intend to do before the next meeting, so it can be corrected now rather than in six weeks.
This costs fifteen minutes to write and reliably transforms the hour. It also changes how you are read: a candidate who arrives with three framed decisions is visibly running their project.
Before the meeting
Send material with a specific request attached. "Draft methods chapter — I would particularly like your view on whether the sampling justification is strong enough for the panel" produces a far more useful read than an unlabelled attachment, because it tells your supervisor where to spend limited attention.
Be honest about the draft's state. Marking something "rough, structure only, please ignore prose" gets you structural feedback instead of line edits on paragraphs you were going to rewrite anyway.
During the meeting
Two habits are worth building deliberately.
Ask for reasoning, not just conclusions. When told to drop a research question, ask what makes it weaker than the others. You are trying to acquire your supervisor's judgement, not just their answers — and in three years you will be the one making these calls unsupervised.
Close the loop explicitly. Before the end, say back what you think was decided: "So I am narrowing to one site, dropping the second research question, and sending you the revised proposal by the 14th." Ninety seconds, and it catches misunderstandings while they are free to fix.
If you have more than one supervisor and they disagree, do not quietly pick one. Say that you have had conflicting advice and ask them to resolve it, ideally in a joint meeting. Candidates who mediate between supervisors rather than making them mediate lose months.
The record, and why you should write it
Write the note yourself and send it within a day. It should be short: decisions made, actions with owners and dates, anything explicitly deferred. Then ask, in one line, for correction if you have misrepresented anything.
This is partly a memory aid. It is also, less comfortably, a protection. Disagreements about what was advised are among the most common serious problems in doctoral study, and the candidate is structurally the more vulnerable party. A contemporaneous record that your supervisor had the chance to correct resolves most of those disputes before they start.
Many institutions require a formal supervision log anyway. Treat it as genuinely useful rather than administrative overhead, and it will pay for itself at least once.
Bring a dated plan to your next meeting A printable phase-by-phase timeline turns a vague progress conversation into a specific one your supervisor can push back on.When it is not working
Sometimes the problem is not process. Meetings that are repeatedly cancelled, feedback that never arrives, advice that reverses without explanation, or a supervisor who has effectively disengaged are real and reasonably common, and they are not something to absorb quietly for another year.
Escalate in order. Raise it directly first, factually and without accusation — "I have not had feedback on the chapter I sent in March, and it is now blocking the analysis" is a statement about the project, not a complaint about a person. If that does not work, most institutions have a second supervisor, an independent adviser, a postgraduate tutor, or a graduate school contact whose role explicitly includes this. Use them. That structure exists because supervision sometimes fails, and using it is not a betrayal.
Changing supervisor is possible and more common than candidates assume, though it is easier earlier than later. The consistent mistake is waiting: candidates routinely spend a year hoping a situation will improve, and the cost of that year is rarely recoverable.