Tips for Academic Job Interviews
It will be interview season for faculty candidates soon, and I thought I would jot down some tips for job interviews in academia/research labs. This is based on attending job interviews in both UT Austin and VMware Research (where I did my post-doc). Please keep in mind that this is all my subjective personal opinion of common problems I saw, and tips based on what has worked for me.
It is always surprising when a candidate who is excellent on paper comes across badly in an interview, and I can’t help feeling with a little more practice, they could done so much better! So this is my attempt to help. This will also be handy to show my own students one day.
First, lets start off with a crucial, high-level point: interviewing well is a different skill than doing great research. You can’t assume that just because you did great research (which is why you got the interview), you’ll have a great interview. You need to practice interviewing.
The job talk is absolutely crucial. It is kind of amazing this has to be said, but we still do get staggeringly bad job talks. The job talk is the single instance when the whole department gets to know you and your research, so you need to come off really well here. A few tips for the job talk:
- Practice, practice, practice. Your first few iterations of the job talk will be garbage. It happens to everyone. Just keep iterating and improving, and you’ll end up with a version that is great. I think I had practiced my job talk over a dozen times before my first academic interview.
- Practice with different audiences. I gave my job talk to my friends, our systems group, grad students in graphics and HCI, a wider systems/databases crowd etc. You really want to reach a wide audience with your job talk, so this is crucial. If only people in your own field understand your talk, you will not get enthusiastic champions from outside your field.
- In my subjective opinion, the best job talks come from HCI and Graphics folks: go talk to your friends in these disciplines on how to give a great talk. When I was at Wisconsin, I talked with Prof. Bilge Mutlu about my job talk, and he gave me amazing tips which helped my job talk stand out.
- Prepare for questions. One good thing about practicing your job talk with different audiences will be getting a feel for what sort of questions you get. When you get a question, you must answer it in a brief manner. Nobody likes candidates who go on for 5 minutes for a single answer: it makes you look as if you don’t know what you are talking about. Anticipate common questions, and have a quick, crisp answer for each one. If you get an unanticipated question, answer as briefly as possible: if they want more detail, they will follow up with another question or ask you later. A back-and-forth looks much better than you droning on for five minutes. This is especially problematic for candidates who want to be thoughtful: they try to look at the question from all angles and provide a thorough, complete answer. IMO, this is the wrong thing to do in a job interview.
- Time yourself, and don’t go over 40–45 minutes without interruption. It is much better to have extra time for questions, than to have a talk that goes over time. Have a plan for what happens if a question took too much for your time, and you have to cut some parts out.
- Don’t try to pack in too much in the talk. This is a classic mistake that a few candidates do every year. You want to give an overview, and go deep into one project. If you try to do deep in three projects in a 1-hour talk, you will come across as shallow instead.
- Make your talk lively! We get a few candidates each year who think it is a good idea to put up equations and explain them in mind-numbing detail. I think the intuition behind most research can be expressed in words; avoid equations. Be as visual as possible. Similarly, use voice modulation, gestures etc. to be an effective speaker; don’t stand behind the podium and talk monotonously.
Research summaries. You need one-minute, five-minute, and ten-minute summaries of your research (I saw this advice first from Steven Hand in a Doctoral Workshop, although I can’t find it). Some schools schedule the talk at the end of day, so you will be meeting professors all day who will keep asking you this again and again. You need to figure out how to explain your research to someone who is not in your field: this takes practice.
Future Work. You need a credible vision of what you want to do for the next few years. Your plan for the next couple of years needs to be somewhat detailed, and after that, it can be a bit hand-wavy. I really struggled with this when I preparing for interview, and eventually got better by talking with my advisors and professors in my area at Wisconsin. Your plan can’t just be “Oh work on something related to Machine Learning”, more like “I want to figure out how to make storage more efficient for machine learning frameworks because..”
See who will you be meeting with. I usually tried to obtain my schedule in advance (not always possible, but you should try) and see who I would be meeting with. Attempt to make a connection (if possible) between your work and the work of whoever you are meeting with. This is much easier to do sitting in a hotel room than in an interview on the spot. Some candidates have sent me follow up emails after the interview; I didn’t do this myself, but I appreciated the dedication of those who did :)
Finally, and this is crucial, express your enthusiasm for joining the department. I cannot stress how important this is. Giving a faculty candidate a job offer is not a trivial thing: many people have to expend political capital to make this happen. They will not do it if they feel you will not be seriously considering the offer. You will not be interviewing at a university if you don’t love it and want to work there: make sure you express this enthusiasm, and if there are factors such as family nearby that makes accepting the offer more likely, please let them know.
Overall, the academic job interview is a grueling process filled with travel and stress. However, it can be also be a way to meet great people in many departments you may not visit otherwise. So enjoy the process! After the first few interviews, most of it will become natural (though still stressful) to you.
Do not take it personally if you do not get an offer: having been on the other side of the table, there is so much politics and strategizing going on that its more or less random as to who gets an offer. It has much more to do with timing and luck than to do with merit; thus, don’t let it go to your head if you do get an offer, and if you don’t, don’t feel too dejected either. I know this is hard advice to follow, but I’m hoping this cushions the blow of any rejects.
Good luck to all the faculty candidates this year!
Great tips from more experienced folks: