The emails, updates, and conversations that keep the relationship working — without burning your time or theirs
Most early-stage PhDs treat their supervisor either as a teacher (asks too much, defers too easily) or as a friend (over-shares, takes critique personally). Neither works. The supervisor relationship is a professional collaboration with asymmetric expertise.
Your job: communicate so they can do their job efficiently. Their job: signpost without rewriting your thesis for you.
Right now, do you tend to under-communicate with your supervisor (avoid them, only contact when you need something) or over-communicate (long emails, frequent updates, every small worry)?
Both extremes are common. The right cadence is somewhere between: a regular, structured update + targeted asks.
Of the three, which one do you find HARDEST to do? Most PhDs struggle with #2 — showing they've tried — because they're worried it'll look like they should have figured it out themselves.
Showing your work is not weakness — it's professional courtesy. It tells the supervisor: "I respect your time enough that I'm not going to ask you to start from zero." It also tells them where your thinking actually is, so they can intervene at the right level. The PhDs who hide their attempts get worse feedback because their supervisors are guessing.
Take an issue you're currently dealing with in your research. Write ONE sentence that describes it specifically (Principle 1). Then write ONE sentence saying what you've already tried (Principle 2).
Your supervisor wants to help. Your job is to make helping you efficient.
Send one of these every 2-4 weeks, even when there's nothing urgent. It builds trust, surfaces problems early, and gives the supervisor a record of your progress without long meetings.
What does the supervisor know after reading this? What do they NOT need to ask you?
They know: you're on track, you're reading, you're aware of what you don't know, and there's one specific thing you want help with. They don't need to ask: how's it going, are you reading enough, is anything blocking you. One short email replaces a 30-minute meeting — and the supervisor will reply faster precisely because the ask is specific.
Done shows you've used the time well — but resist listing every small thing. Three or four meaningful items.
In progress tells them where you currently are and lets them intervene early if the direction is wrong.
Stuck is the heart of the email. Frame the stuck specifically. Show what you've tried.
Next two weeks commits you publicly. It also gives the supervisor a chance to redirect before you spend two weeks on the wrong thing.
Use the four-section structure. Write the email you would send today.
Look at your "Stuck" section. Is it specific enough that your supervisor could give you a useful pointer in a single sentence? If not, rewrite it until it is.
Every 2-4 weeks. Same structure every time. Predictability is a feature, not a limitation — your supervisor scans these in 90 seconds.
The most common supervisor email — and the one PhDs get wrong most often. The error is asking for feedback on too much, with too little direction, and no deadline.
Three problems: no scope (the whole chapter), no specific concerns flagged, no deadline. The supervisor will defer it indefinitely — and probably skim when they finally read it.
Scope is bounded (one section). Two specific questions. Type of feedback specified. Deadline proposed but flexibly.
What's the difference between asking for "structural feedback" and asking for "line-level edits" — and why does it matter?
Supervisors waste enormous time correcting commas in chapters that will be restructured. If you tell them what kind of feedback you want, they give you better feedback faster. Structural early, line-level late. Saying "structural only at this stage" tells them you know what you're doing.
Pick a piece of writing you have or are about to send. Draft the email. Specify scope, two questions, the type of feedback, and a flexible deadline.
The more bounded your ask, the better the feedback. A four-page chapter with one specific question gets better feedback than a forty-page chapter with no question at all.
It will happen. They will recommend something you think is wrong, or push back on a position you're confident in. The instinct is either to capitulate (accept everything to keep the peace) or to argue defensively. Neither serves the thesis.
The professional move is to disagree in writing, with reasons, and offer them an easy way out.
If you accept every supervisor suggestion uncritically, what happens to your thesis by viva day?
You can't defend it. The viva panel will ask you why you made certain choices — and the answer "my supervisor told me to" is fatal. The thesis has to be defensibly YOURS. Disagreement, handled properly, is part of the apprenticeship.
What three moves does this email make that turn disagreement from confrontation into collaboration?
Imagine your supervisor recommended a methodological change you don't agree with. Write the reply — using the structure above.
You can disagree and stay collaborative. The signal is reason + humility + invitation to talk. Avoid sarcasm, avoid victimhood, avoid the passive voice. Be direct and warm at the same time.
Some supervisor situations need careful handling. Here are the four you're most likely to face — and the language for each.
You sent a chapter draft three weeks ago. No reply. Email three? Or wait?
Send a polite, low-pressure follow-up. The trick is to NOT make them feel guilty (which makes them avoid replying further) and to give them an easy way out.
You receive feedback that feels brutal. Your instinct is to write back defending the work. Should you?
No. Wait 24 hours. Then read the feedback again. Then write a calm, professional response that thanks them, asks for clarification on one or two specific points, and proposes next steps. Defending in heat damages the relationship and obscures whether the feedback was right.
If you have co-supervisors and they give you contradictory advice, do not pick a side privately. The professional move is to name the contradiction in writing, copy both supervisors, and ask them to agree a position.
What is the email NOT doing? What is it carefully avoiding?
It's not taking sides. It's not implying either supervisor was wrong. It's framing the contradiction as a coordination problem (their problem) rather than a judgment problem (yours). It also offers them an out — "one of you take the lead" — which they may accept.
You're three weeks behind on a milestone. Do you (a) work harder and try to catch up silently, or (b) email your supervisor to flag it?
(b), every time. Supervisors hate being surprised by missed milestones at month 18. They are fine with delays they've been informed about — and often have practical solutions you don't have access to (extension funding, rescheduled meetings, etc.). Silence is the worst option.
The supervisor relationship works best when you treat your supervisor as a busy, intelligent collaborator — not as a teacher, friend, or judge. Brief, structured, specific, professional. Same as any working relationship that needs to last three or four years.
Three actions before next lesson:
Bring all three to next lesson. We'll pressure-test the wording.