How to Find a PhD Supervisor Before You Apply

How to Find a PhD Supervisor

Work backwards from recent papers, not rankings. Read two years of output on the problem you care about, note who is still publishing on it, then check who is taking students this cycle. That list becomes your program list.

Why the paper trail beats the faculty directory

Directory pages describe interests as they were when the page was written, often years ago. Recent papers, lab news and current group listings tell you what someone is working on now, whether the group is funded, and whether there is room in it. Rankings tell you none of that, and a department can be excellent in the abstract and empty of anyone working on your problem.

Finding is also a different job from emailing. The University of Washington's Allen School tells doctoral applicants that "You do not need to contact faculty prior to applying, nor is it expected", while adding that an applicant confident about a particular faculty member "may indicate their name(s) in the supplemental question." Where admission runs through a committee, the search pays off inside your application rather than in anyone's inbox. Where it runs through an advisor match — much of psychology, and most positions funded from a single grant — the same search decides whether there is a match to be made at all. Either way the work is identical, and whether to write to anyone is a separate question.

A search that produces usable material

1. Start from a problem, not a department. Take the last two years of the main venues in your subfield and note every group that published on your problem.
2. Read the author list downward. The senior name may be running a large group; the person doing the day-to-day work is often a newer faculty member with funding and space.
3. Check they are active. Recent publications, current students listed, a lab news page, any admissions note saying whether they are recruiting.
4. Record one specific thing each. A method, a dataset, a result you would want to extend. That sentence is what ends up in your statement.

The fix: two or three named, evidenced pairings

A supervisor search has done its job when you can write a sentence nobody else could write about that group.

Illustrative excerpt — written by GradPilot for this page, not a real applicant's statement.

Weaker: "Professor Alvarez's research aligns closely with my own interests in graph learning, and it would be a privilege to work in her group."

Stronger: "Professor Alvarez's group reports that its message-passing model degrades on sparse citation graphs. I spent last summer building sparse-graph benchmarks for a course project, and that failure case is what I want to work on."

The second names a problem inside the work rather than a topic label, and attaches something the applicant has actually done to it. Two or three of those beat a list of eight names.

For outreach timing, templates, follow-ups and what faculty say about cold email, the long-form companion is how to find and contact a PhD supervisor.

Check your own draft

Take every faculty name in your statement and ask whether the sentence around it could be pasted into another application with one word changed. If it could, the search is not finished — you have a name but no pairing.

For how fit evidence differs across fields, see the SOP examples by field hub. The published PhD adviser fit rubric shows what a reader is looking for when you claim a match, before you submit anything.

Two free Quick Reviews a day come with every account: run the draft through one and see whether your pairings read as evidenced or as flattery.