How to Manage PhD Students Without Chasing Them

To manage PhD students without chasing them, give the week a rhythm that runs itself: every student writes a three-minute update on Friday (what they did, what's next, what's blocking them), supervisors get a digest on Monday with the blockers at the top, every 1:1 keeps notes whose lines become tasks, and degree milestones remind everyone 30, 7 and 1 days ahead. The PI then works from one list — who needs me this week — instead of from memory, email and hallway conversations.

Why supervision turns into chasing

A professor with eight PhD students, two postdocs and a few Master's students is running a small organisation on top of teaching, grants and reviewing. Without a rhythm, the only way to know how someone is doing is to ask — and asking scales badly. Every question is an interruption for both sides, the answers arrive in different places, and nothing adds up to a picture.

The costly failures are the quiet ones. The student who has been stuck on a cluster permission for three weeks doesn't announce it; they try another workaround. The first-year who is unhappy doesn't say so in a group meeting. The qualifying exam that should have been scheduled last month isn't anyone's job until it's late. By the time a supervisor notices, the fix is bigger than it needed to be.

The answer is not more meetings. It is making the information arrive on its own, in the same shape every week, where the supervisor already looks.

1. The Friday update: three minutes, four questions

Keep it short enough that nobody resents writing it and nobody dreads reading it:

• Done this week — experiments run, sections drafted, papers read, results that surprised them.
• Next week — what they'll do, in a line or two.
• Blockers / need help — the most important box. Anything stuck goes here: a dataset access request, a bug, a disagreement about direction, a reviewer comment they don't understand.
• Reading (optional) — papers worth sharing with the group.

An optional "How are you doing?" from 1 to 5 is worth adding, but only if it is visible to the student's own supervisors and nobody else. Students answer honestly when they know peers can't see it. A score that drops two weeks in a row is often the earliest signal a PI gets that something outside the lab is going on.

Timing matters more than format. A reminder on Friday afternoon, in the student's own time zone, gets most updates written while the week is fresh. A gentle nudge on Monday catches the rest — sent automatically, so the supervisor never has to be the person asking.

2. The Monday digest: blockers first

On Monday morning each supervisor gets one summary: how many updates came in, who is blocked, and who hasn't written. The PI gets the same for the whole group. Reading eight updates takes ten minutes when they arrive in one place with the blockers on top; it takes an hour of inbox archaeology when they don't.

Two habits make the digest work:

1. Answer every blocker the same day. Even "let's talk Wednesday" tells the student the flag was seen. Blockers that get no response stop being written.
2. Nudge, don't chase. A one-click "please write your update" is enough; the rhythm carries the rest. If someone misses three weeks running, that's a conversation for the 1:1, not a reminder.

An AI summary of the week can help a PI with a large group see themes — three students blocked on the same GPU queue, two waiting on the same collaborator — as long as it only restates what the updates say, never invents results, and falls back to the plain list when it is unsure. Wellbeing scores should never leave the database, AI or not.

3. 1:1s with a memory

Most 1:1s start with "so, where were we?" A better agenda builds itself from three things:

• the blocker from the student's last update,
• the action items still open from the previous 1:1,
• the milestones coming up in the next month or two.

During the meeting, write notes one point per line. Any line that needs doing becomes a task with an owner and a due date — "rerun experiment 3 with the new seed, Priya, Friday" — and stays on the next agenda until it's done. That single habit turns 1:1s from conversations into a record, and it protects students too: what was agreed is written down in a place both people can see.

Track when each student last had a 1:1. If the group agrees on "every two weeks", a student who hasn't had one in three should show up on the PI's list — not be discovered in month four. Booking the next one should be one click from the notes, with both calendars' busy times visible (the same idea as booking internal meetings without the back-and-forth).

4. Milestones that remind themselves

A PhD has a shape: coursework, a qualifying exam, a thesis proposal or candidacy, annual reviews, papers, a thesis draft, submission, the defense and corrections. A Master's or undergraduate thesis has a shorter one. Lay the milestones out once from a template for the program and the student's start date, adjust them to the person, and let reminders go to the student and their supervisors 30, 7 and 1 days before each one.

Three details make milestones useful rather than decorative:

• A status the student can set — planned, on track, at risk — so "at risk" is something they can say before it's late.
• A supervisor's sign-off, recorded with who and when, so "done" means done for the annual report.
• Links, not attachments — the Overleaf project, the committee form, the draft — so the milestone points at the work.

A group timeline — every student's start, expected finish and milestones on one chart — answers "who graduates when?", which is the question behind hiring, funding and teaching plans.

5. One list: who needs me this week

Put it all together and the PI's home page becomes a single list, ordered by urgency:

1. Students who are blocked
2. Approvals waiting — leave, purchases, equipment bookings
3. Missing weekly updates
4. Milestones and paper deadlines this month
5. Overdue 1:1s
6. Funding or visa end dates within 90 days
7. Recommendation letters due

Each line opens the thing that needs doing. On a quiet week the list is short, and that is the point: the absence of items is information too. Students get their own version — their update status, their next 1:1, their open action items, their milestones.

6. Privacy is part of the design

Weekly updates, 1:1 notes and wellbeing scores are sensitive. The rule that works: a student's records are visible to the student, their own supervisors and co-supervisors, and the PI — not to peers, and not to lab managers who run equipment and orders. The same boundary should hold for funding and visa dates. If the tool can't enforce that in the database, students will (rightly) write less. For how that kind of scoping works in practice, see how leads see their team's work without seeing everyone's.

How LetRelay does it

LetRelay has a Research group / lab workspace type that switches all of this on: weekly updates with Friday reminders and a Monday digest, 1:1 notes that turn into tasks, milestone templates with supervisor sign-off, a paper pipeline with AoE deadlines, equipment booking with training rules, grants, recommendation letters and printable progress reports — plus a "This week" home for the PI. Visibility is enforced with row-level security in the database, not just hidden in the interface. A small lab can run it on the free plan; the full list is on the LetRelay for research groups page.

FAQ

How often should PhD students write updates?

Weekly works for most groups: short enough to stay current, long enough that each update has substance. Keep it to three or four questions so it takes minutes, and make the "blockers" question the one supervisors always answer.

Should supervisors see a student's wellbeing score?

Only their own supervisors and the PI, and only if the student chooses to answer. It should be optional, never visible to peers, and never sent to third-party AI tools.

How often should a PhD supervisor hold 1:1s?

Every one to two weeks is common in lab-based groups; monthly can work for senior students writing up. What matters more is consistency and a written record of what was agreed.

What milestones should a PhD plan include?

Typically coursework, a qualifying exam, a proposal or candidacy, annual progress reviews, papers, a thesis draft, submission, the defense and corrections. Adjust the list to your program's regulations.

Sources

• CRediT — Contributor Roles Taxonomy (NISO)
• Crossref REST API documentation
• arXiv API user's manual
• IANA — Time zone database