How often should you meet your PhD scholars?

There is no universal right answer, but there is a wrong way to arrive at one: picking a cadence when the scholar starts and never revisiting it. A PhD runs four to six years, and the supervisory relationship at month four has almost nothing in common with the one at month forty. A fixed cadence is right for at most one of them.

Start from what goes wrong at each extreme

The failure modes are not symmetric, which matters more than most cadence advice admits.

Meeting too rarely fails slowly and expensively. A scholar working in the wrong direction does not stop being productive — they produce a great deal, all of it pointing somewhere you would not have sent them. The cost is measured in months and lands almost entirely on the scholar. This is the failure that produces the fifth-year crisis where a chapter has to be abandoned.

Meeting too often fails quickly and visibly. The scholar has nothing new to report, the meeting turns performative, and both parties start treating the slot as an obligation. The real damage is subtler: a scholar who meets their supervisor weekly for four years may never develop the tolerance for being stuck that research requires. Some of the most important work in a doctorate happens in the fortnight where nothing works and nobody is coming to rescue you.

The asymmetry matters. Too rare is worse, because it fails silently. If you are unsure, err toward more contact — but notice that more contact need not mean more full-length supervision meetings.

A defensible default by stage

First six months: weekly, and keep them short. A new scholar does not yet know what a research question looks like, how much reading is enough, or what standard of evidence their field expects. None of that can be taught in a document. It is absorbed through frequent, low-stakes contact. Thirty minutes weekly beats two hours monthly at this stage by a wide margin.

Months six to eighteen: fortnightly. The scholar is running their first real piece of work. They need enough room to get stuck and unstick themselves, and enough contact that a wrong turn is caught within a fortnight rather than a term. This is where the agenda should shift from what did you read to what did you find and what do you think it means.

Middle years: fortnightly to monthly, driven by the work. Once a scholar is executing a plan you both believe in, cadence should follow the work rather than the calendar. Longer gaps during a long data collection run; tighter contact around analysis, writing, and anything going to a committee. The one thing not to do is let the gap drift without deciding to. A cadence that has quietly become quarterly because you were both busy is the most common way a scholar goes missing.

Final year: it depends entirely, and you should say which. Writing-up scholars divide sharply. Some need frequent short check-ins to keep momentum against a task with no natural structure. Others need long uninterrupted stretches and find meetings actively disruptive. Ask which they are. They usually know, and almost nobody asks.

Separate the kinds of contact

Most of the tension around cadence disappears once you stop treating all contact as the same thing. There are at least three kinds, with different natural frequencies.

The substantive supervision meeting — one topic, thought about properly — is expensive, and does not benefit from being weekly for an experienced scholar.

The check-in is five minutes and answers one question: is anything blocking you that I could unblock? It costs almost nothing and catches the failure above, where a scholar has been stuck for three weeks on something you could have resolved in a sentence. A weekly check-in plus a monthly substantive meeting is a far better arrangement than a fortnightly meeting trying to be both.

Written contact — comments on a draft, a reply to a question — scales in a way meetings do not. A scholar who can get a useful answer in writing within two days needs fewer meetings than one who cannot.

Write the cadence down, and set a date to revisit it

Whatever you choose, make it explicit, and put a date on when you will reconsider. "We will meet fortnightly and review that in six months" is a completely different arrangement from a fortnightly pattern nobody ever agreed to and nobody can therefore change without it feeling like a complaint.

This matters most for the scholar. A scholar who thinks they are getting less contact than they need has two options: raise it, which feels like criticising their supervisor, or say nothing. Most say nothing. An agreed review date turns an awkward complaint into a scheduled agenda item, which is the entire point.

The number that actually predicts trouble

If you track one thing, do not track average frequency. Track the longest gap.

Averages hide exactly the failure you care about. A scholar you met eight times this year looks fine on an average until you notice six of those meetings were in October and there is a fourteen-week hole between February and May. That hole is where scholars get lost, and it is nearly always invisible to the supervisor, who remembers the meetings rather than the silences.

The gap is also the easiest thing to fix once you can see it. It usually is not a decision anyone made. It is a cancelled meeting that was never rescheduled, twice in a row, during a busy term.