A campus visit can run two full days, put you in front of thirty people, and come down to the forty minutes of questions that follow a talk you spent a month building. Interview advice written for a one-hour conversation with a hiring manager and a recruiter does not transfer to this, and candidates who prepare that way lose to weaker candidates who understood the format. The format is the thing worth learning.
The job talk carries the decision
The committee is not only asking whether your research is good. Two quieter questions run underneath the whole talk: would we want to sit in your seminars for the next ten years, and is there work here once the current project ends.
The second one catches people out. A talk that closes on a finished, polished result answers the question "did this person do something good", which nobody doubts by that point. A talk that closes on the next two studies, the collaborators they need and where the money for them comes from answers a different question, and that is the one being scored.
Pitch for the room you actually get. Most of the people listening are not in your subfield and a few of them are not in your discipline. A colleague from two corridors away should be able to state your problem and why it matters after eight minutes, without having understood a single method slide. Put the general framing at the front, the technical core in the middle, and the programme at the end. Ask the organiser whether the advertised slot includes questions, because half of them do and half of them do not, and build to roughly two thirds of the time you are given.
The question period is where it is won or lost
You get four broad kinds of question. The genuinely curious one. The local-interest one, meaning "could this work on the data we already have". The testing one, from someone who wants to see you under pressure. And the statement dressed as a question, from a senior person with a position of their own.
That last one is the trap. Acknowledge the position, say where you agree with it, then give your own view and the reason for it. Folding makes you look like someone who will not defend a grant application. Fighting makes you look like a difficult colleague. "I don't know" is survivable when followed by how you would find out. Bluffing is not, because someone in that room knows.
The committee is not scoring your best result. It is working out whether the next decade of your work is something worth building a department around.
The teaching slot is not a short job talk
This is where candidates from research-heavy backgrounds give away ground they cannot recover. The brief is usually something like "teach thirty minutes to second-year undergraduates on this topic". A surprising number of people respond with a compressed version of their research interests, which demonstrates nothing about instruction.
Ask questions before the day. Which module does this sit in, what have the students already covered, will there be real students in the room or staff playing the part, and does anyone assess it. Then teach: one idea, a worked example, a question thrown to the room, a visible check on whether it landed. Use the board if you would normally use the board.
If real students are present, teach them and not the panel. Take a wrong answer well. People remember how you handled the student who got it wrong long after they have forgotten your slides.
A written teaching statement is often requested alongside. Concrete beats philosophical every time: what you assess and why, what you do with a cohort where a third have the prerequisite and the rest do not, how a lecture-hall course differs from a twelve-person seminar in your hands specifically.
Everything else on the schedule is also the interview
Lunch with graduate students. The walk across campus. Twenty minutes with the dean. Coffee with someone whose work has no connection to yours. None of it is a break, and all of it reports back, usually within the hour. Graduate students in particular get asked directly for their view, and their view carries weight because they saw the version of you that had stopped performing.
The meeting with the dean or the head of school is a resources conversation wearing a friendly hat. Have a specific answer ready for what you need to do your work there: lab space of a particular kind, reduced teaching in the first year, one funded doctoral position, a named piece of equipment. "Just some support" reads as no plan. A wish list with no sense of the budget reads as someone who has never run one.
Practical and slightly undignified: eat something, and go easy at the dinner. Two days is a long time to stay sharp on coffee and nerves.
The questions you only get in this setting
Funding over the next five years. They want calls with names and dates, who you would apply with, what you do if the large one fails, and what a small internal award would buy you in the meantime. The schemes and the calendars differ by country and sometimes by institution type, so speak about your own system rather than a generic one, and do not cite a call you cannot discuss in detail.
How you fit the existing strengths and the gaps. Departments hire against a hole they have identified, and you are being measured against that hole. Read the staff list, the module handbook and the recent thesis titles. Name two people whose work genuinely touches yours and say what the joint output would be.
What you would teach outside your specialism. Core courses have to be staffed, and a candidate who can only teach their own research is an expensive luxury. Answer with two modules from their own handbook you could take over next year, plus one you would propose.
Supervision. How many doctoral students at once, what you do with one who has stalled for six months, where you stand on authorship, how you handle a project that is drifting away from what was funded.
Fit is a real criterion here, not polite filler
In corporate hiring, fit is often vague. In an academic appointment it is substantive, and it is contested, because the committee members want different things. One person wants an ally in their subfield. The head of school wants the Tuesday morning course covered. The dean wants research income. The graduate students want a supervisor who answers email. Generic multi-interviewer technique helps a little here, but the real work is choosing a coherent story and holding it, because a candidate who reshapes their pitch for each listener ends the day with no identifiable position at all.
Running this while you hold another post
The timeline is long. Months pass between application and offer, and a single search can occupy a full academic year. Budget leave days, because two-day visits are not something you can slot into a lunch break.
References work differently here. In many academic systems referees are contacted before shortlisting rather than after an offer, so tell your referees early, send them the advertisement and your materials, and warn them when a visit is scheduled. If you are also running non-academic applications in parallel, keep the documents separate; something like Postulit can build a conventional CV from your LinkedIn profile so your academic one stays what it needs to be.
Negotiation comes after the committee makes its recommendation, and the salary is often fixed on a published scale with little movement. The start date, the start-up package, first-year teaching relief and space usually have more give than the number does.
Start with the eight-minute version
This week, write the eight-minute version of your talk: the problem, why it matters, what you did, what comes next. Deliver it to someone outside your field and ask them to describe your research programme back to you. If they cannot, the committee will not either. Then write down the four questions you least want to be asked and answer them out loud, standing up, until the answers stop wobbling.