Tips for a Massive Academic Job Search

Ellen Spertus

[I wrote this essay in spring 1997.  I updated it in February 2009, making minor edits, some of which are shown in brackets; removing, replacing, or fixing broken links; and adding an afterword.]


I can't believe that finishing my dissertation will be nearly as challenging as applying for over a hundred academic and research positions, spending seven weeks on the road, juggling offers with different expiration dates, solving the two-body problem, and negotiating a contract, which is how I've spent the spring of 1997. This document is meant as advice for other graduate students based on my experience and stories I've heard. I've made an effort to point to and not duplicate information available elsewhere. The best exhaustive site on the PhD experience is

The primary purpose of a job search is to find the best possible job. This can be divided into two components: determining what your ideal job is and convincing others to give you that position. When I began my search, I didn't know if I wanted to be at a research lab, a research university, or a liberal arts college. I wasn't even sure what academic department I was most interested in. My job search was as much about finding out what I wanted as convincing others to hire me.

This document will be most useful to female computer scientists whose partner is in computer science and also looking for a job, although most of it should be relevant to any doctoral candidate seeking a job. Parts of it will be relevant to search committee chairs wishing to be hospitable to job-seekers.

The Early Stages

Your job search begins when you get to college. Not only is that when you being to network, but it's also when you begin to learn what it's like to be a professor. Networking is covered in other documents, such as Phil Agre's Networking on the Network. It's harder to find advice about how to determine what type of job you want. Because you spend so much time in graduate school, it's easy to learn what life is like for professors at research universities. I don't think you can intelligently compare job categories unless you've also worked for at least a summer in industry (probably as an undergraduate) and at a research lab (probably as a graduate student). Unless you're certain you don't want to teach, you should get teaching experience. This will both help you determine whether you like teaching and make you a stronger candidate for teaching positions.

Another way to find out about the trade-offs of different jobs is to observe the people around you. In your years at graduate school, you'll see professors climbing the tenure ladder and ones who have made it. Learn from them. I observed that junior faculty members tended to get divorced, turn prematurely gray, or lecture classes right up until they went into labor. I also watched the choices made by students who graduated before me. Whatever the choice, you can learn interesting things by asking the person how he or she came to it and staying in touch so you can see how they feel after a few years. Also see John Wilkes' Is work hell? Life in industrial research.


If you will be applying for many jobs, organization is crucial. My major organizational components were a spreadsheet for keeping track of applications, a word processor with support for address labels, a wall map for keeping track of geography, and a filing cabinet for application materials.

Whenever I found out about a position, I created a new row in a spreadsheet. [I used Microsoft Excel.  I'd advise someone nowadays to use an online spreadsheet, such as Google Docs.]  The column headings included the school's name, city, state, number of positions, deadlines, the job listing URL, the department's URL, required application materials, whether letters were supposed to be sent directly, and the status of my and Keith's applications, including when we applied, what we sent, and when we received an acknowledgment. (Additional columns will be described later.) This allowed me to easily view, for example, all employers with deadlines of January 1 to which I had not yet applied or all schools that I'd applied to but had not yet received an acknowledgment from. The spreadsheet contained all the schools that either Keith or I applied to, so we were able to share the file. We also shared it with other graduate students. In a two-body job search, efficiency can be greatly increased by having each person work on what he or she is best at or prefers. For example, I did most of the typing, and Keith did most of the photocopying.

The word processor I used supported address labels, allowing me to type an address once and use it many times: on my letter, on Keith's letter, on our mailing labels, on mailing labels for our references, etc. Even when it's more work, you should cut and paste an address instead of retyping it, to avoid introducing typographical errors. You should create a default cover letter and then personalize it carefully for every application. I sent out at least one letter expressing my interest in "the Department of <<FILL IN >>". You should save an online copy of every letter you send out, in case you need to resend or review it.

In order to keep track of geographical constraints, I bought and posted a road map of the United States and southern Canada, putting different colored map pins wherever we were applying. When one of us received an invitation to interview, I replaced the pin with a thumbtack so it was easy to see the status of our job search visually. The map also helped in planning the order of visits. Also useful was the web service How far is it?.

I was able to fit all of my job search in one hanging folder, which held a regular manila folder for copies of each component of my application packet, i.e., my CV, teaching statement, research statement, transcripts, and reprints of various papers. Other things to keep on file are stationery and envelopes, which you'll run through much faster than you expect. When you make copies of some component, like your CV, I advise you to make many. Keeping the folders full makes the marginal cost of an application small: just grab something from each folder and pop it in an envelope. You will invariably waste paper when you have to update material, but I don't think this can't be helped. Plant a tree when you graduate. Don't try to economize by using both sides of a piece of paper. I sent double-sided CVs to a few places I applied, and, at two places I visited, I discovered that people I met with had only half my CV, since only one side had been copied. For this reason, you should also number your pages.  [Nowadays, many, but not all, schools accept online applications.   One reason some schools require paper applications is because fewer people apply, due to the effort, saving the search committee time by eliminating less interest applicants.]

Determining Where to Apply

I followed my advisor's advice and applied anywhere that I thought it was even remotely possible that I would be interested. She told me that she sometimes hadn't realized how attractive a place was until after she reluctantly accepted their invitation to interview. Another professor advised me to instead select no more than ten places that I was most interested in and to really focus on those. I combined the approaches by applying everywhere and putting in extra effort to the likeliest places.

You can find CS faculty ads posted at the CRA, ACM, and IEEE sites and in the corresponding print magazines. You can also get on mailing lists, such as the CRA job list. There are sites that charge for access to the newest job listings and provide older ones for free, such as  The Chronicle of Higher Education. I found a six-month subscription to the Chronicle to be well worth the cost ($40.50). Not only could I electronically search the latest job listings, but I could easily research a school by searching the archives and could access data about professors' salaries, etc. If you're interested in teaching-oriented schools, in addition to reading the Chronicle, you should also join SIGCSE to get access to its job mailing list, or find someone to forward you the messages. I advocate sharing information with other graduate students. While it's possible that the job you want won't be offered to you because you told someone else about it and they applied, it's also possible that the job you want will be available because someone turned it down to accept a position you told them about. Besides, it's good karma.

One reason to eliminate jobs at this stage is a two-body constraint (a partner who is also seeking a job), unless you're willing to share a position. Keith and I only applied to places that had multiple positions or were near other employers. (Actually, before we were engaged, I applied to a few places in the boondocks. When Keith observed that there were no jobs for him nearby, I said that I didn't know yet what his intentions were.) He agreed to apply for jobs that weren't at the top of his list, and I did the same, to maximize the chances of finding a two-body solution. Neither of us played any games with only applying to places high on our list to try to force the other to make any sacrifices. If you don't want what's best for both of you, you shouldn't be getting married.

Keith wanted to work at a research-intensive institution, so he eliminated schools that didn't have PhD programs or had poorly-rated research programs. The Computing Research Association has a very useful tool for ranking PhD-granting CS departments [although the data is now very out-of-date]. I wanted to work with bright students, whether or not they were graduate students, so I eliminated schools that were rated as being less competitive, based on their undergraduate admissions criteria. You can buy a college guide (often at a used book store) to find out what tier a school is in, or you can find the information online through U.S. News and World Report. We added some of this information to our spreadsheet. You might also want to look for schools in a certain category, such as historically-black colleges and universities (HBCUs) or science-active liberal arts colleges, and guides are available for these. The best qualitative general guide to colleges I found was The Insider's Guide to the Colleges by the Yale Daily News staff. Take what you read in these books with a grain of salt, however. I read in Lisa Birnbach's College Book that Mills students wear "severe daytime dress-up... and have the luster of a newly lacquered manicure." Not when I visited. All of the students I saw were dressed casually, and they had just constructed a giant mud ball.  [Nowadays, I would recommend the Princeton Review website.]

Some people apply to departments that haven't advertised positions. I didn't try doing this, although I did email some department heads at interesting schools to ask if they anticipated any openings; I didn't find any this way. When I applied to multiple departments at one school, I wrote in each cover letter that I was applying to both and would consider a joint position.

The Application

Ask for the job packets of successful job-seekers in order to create first drafts. Then, have your advisor and any other friendly professors suggest changes, which you should take very seriously. I was able to get very useful advice from MIT professors on applying to research universities. Since I was also interested in liberal arts colleges for women, I put together a draft job packet well before any were due and sent it for criticism to a friend who is a senior professor at such a school. I also found much useful information in Malcolm Campbell's How to Get a Teaching Job at a Primarily Undergraduate Institution.

Make every effort to send in an application by the requested date. While some schools accept late applications without complaint, there are other schools that return or discard applications postmarked after the deadline, as I learned the hard way. You can always update an application by sending further information later. It is your responsibility to make sure that letters of recommendation arrive, especially if the position announcement says that you should arrange for the letters to be sent (as opposed to their soliciting them if your file seems interesting). You should find out from each reference how he or she likes to receive requests: through an assistant, with printed labels, with online labels in various formats, etc. You should send requests for multiple places together, instead of dribbling them in one at a time. Many places won't consider your application until it is complete (i.e., all letters have arrived) and don't tell you if it's incomplete. It is your responsibility to contact the professor or administrator in charge of the search to find out which letters have made it into your folder. Don't assume that because a letter got sent that it arrived to the right place. Many letters go astray and have to be resent.

Most schools will send you an acknowledgment within a few weeks of receiving your application or the deadline. If you do not get an acknowledgment, email or telephone the contact person. (I did this and discovered that something had indeed got lost along the way to one school.) You will also receive an affirmative action form from most places, requesting information on your race, sex, and how you found out about the position. (For this reason, you might want to include a column on your spreadsheet indicating where you first found out about a job.) Returning the forms is not required, and they are not part of your application. They go not to the search committee but to the affirmative action office, so they can document what type of people applied for the job and that it was properly advertised. Don't assume that if you're white that you're at a disadvantage. According to the 1996 Taulbee Report, "a much larger percentage of white non-Hispanics are hired into tenure-track positions than are awarded PhDs." Similarly, I have seen little advantage for female candidates – and more personal questions, which I'll get to in the section on interviewing.

For every position you care about, you should also find a personal contact. This is not optional. If you know a professor in the department, contact him or her and say that you're applying for a position and that you're very interested in the department. If you don't know somebody, ask your references, especially your advisor, to put in a good word for you, or see if you can find someone in a professional organization you belong to, such as NOGLSTP or Systers. I cannot stress enough how important a personal contact is. There was one school I was interested in – let's call it Stateville – because I had heard good things about it from a relative who had attended and because it was located in an area I wanted to live. I was very highly-qualified for the position and expected to be invited for an interview as soon as they started the process. I wasn't. I later found out this was because they thought I was out of their league, especially because they'd heard I had been invited to a more prestigious school in the area, which we'll call Joliet. I actually would have preferred Stateville to the more prestigious schools. By the time I found out what was going on from someone I met at Joliet, had her communicate my interest, and got an invitation, it was too late in the season for me to seriously consider them. Avoid these situations by making use of social networks early.

Responding to Invitations


Typically, a professor will call you (or ask you to call) to extend an invitation. A few financially-poorer schools wanted telephone interviews first. It was at this point that Keith and I mentioned the two-body problem, if there was no other employer in the area, although some couples don't say anything until one receives an offer. A couple that went through the process before me said that mentioning their spouse at the invitation stage usually got the spouse an invitation. That wasn't the case for us. We got invited to interview at two of the same schools, but that appears to have been coincidental. A few liberal arts colleges asked if we would be interested in sharing a position; we weren't. (While sharing positions helps some people solve the two-body problem and to balance an academic career with an outside life, my impression is that it's more than half-time work for each person at only half pay.) When told about the partner, we were sometimes told about job possibilities in the area and in one case received the suggestion that the other person apply for a research scientist position. Schools still encouraged us to fly out to interview even after we pointed out that our two-body constraint seemed unsolvable with them. Keith and I both had the experience of being invited to a school that wasn't interested in the other person. It might be hard psychologically if one partner is universally seen as the stronger candidate.

Some places offered more than a month's notice and a choice of interview dates; others said, "Be here by Wednesday or forget it." People typically have their first interview be at a place that's not their top choice, since they'll get better at interviewing as the season progresses. In general, however, I'd recommend interviewing at the places you're most interested in early in the season, particularly if you have a two-body problem. Keith and I timed things poorly, and I received an attractive Bay Area offer (with a two-week deadline) a week before Keith's first Bay Area interview. Another reason to schedule less attractive schools later is that you can cancel them if you get a more attractive offer before visiting. This is not rude. I've always been thanked when I've told a host that because of an offer I've received that it no longer makes sense for me to interview with them.

Because the invitations will arrive in a random order, possibly after you've started traveling, an optimal ordering is impossible. I thought of it as an on-line (on-the-fly) traveling salesman problem or sometimes as a treasure hunt, where in each city I'd find out what the next city would be. While you can't do an optimal scheduling, there are things you can do to make your scheduling easier:

The most strenuous weeks I booked had me doing a two-day interview on Monday and Tuesday, resting on Wednesday, and interviewing at two separate places on Thursday and Friday. Scheduling anything more intense than this is not recommended. After two days on, you should have a day off if at all possible. You can spend a rest day traveling if it's a non-stop flight, but it won't be a restful day if you're having to change planes and worry about connections. I'd recommend scheduling no more than three days of interviews and one day of traveling in a week. One two-day interview is less strenuous than two one-day interviews, since (1) you'll only have to give one talk (generally), (2) the second day is sometimes shorter, (3) you'll be going back to a familiar place, and (4) you won't have to do any traveling in between. You can make life easier for yourself by requesting at least 15 minutes to your self immediately before your talk for preparation (or just for resting your head on a desk).

If you have any special needs, now's the time to make arrangements, such as those made by Elizabeth Lane Lawley, who became an assistant professor at Rochester Institute of Technology:

I brought my (then 5-month-old) baby with me on interviewing trips. I was still breastfeeding, and was unwilling to wean for the purpose of a job hunt. So I arranged for child care in advance in each city, but also told the search chairs that I'd need private time at lunch each day in order to express milk for him. I figured that if they still offered me a job, I'd know they were basically mother- and family-friendly.

Air Travel

You should belong to the major frequent flier programs. To consolidate your miles, find out which airlines are partners.  Some air travel tips:

Preparing for Interviews

The most strenuous part of the job search is traveling the interview circuit. The more you prepare in advance, the less stressful and more productive your trip will be. As I mentioned, I was on the road for over seven weeks, making it back to my home base (Seattle) for only two (separate) days during this time. You can get advice elsewhere on how to prepare a talk, what questions to expect, and so forth. You'll get the best job talk advice from people at your own institution who attend your practice talks. Be sure to include some people who aren't already familiar with your work.

The first thing you need to prepare for interview travel is money. You'll have to pay for dressy clothing, accessories, and travel expenses, such as airfare. You'll eventually get reimbursed for most of the travel expenses, but it won't happen overnight. I fronted over $3000. If you don't have the money, borrow it. Some advisors have been willing to make bridge loans to students.

You can save money on clothing by starting to shop for it well in advance, keeping an eye out for sales. Consider also checking out consignment shops (high-quality used clothing stores). You should have at least two dressy outfits, because some interviews last two days. You will also need some non-grubby comfortable clothes, because a faculty member may pick you up at the airport and take you straight to dinner with other people from the department. Make sure that your clothing is practical and doesn't wrinkle too easily. Women: When you're tempted to buy a dress or skirt that is knee-length or shorter, ask yourself whether you'll feel comfortable in it sitting next to someone, climbing up stairs, getting out of a car, etc. Also, make sure your shoes are comfortable, because you will have to walk around campuses. I find Easy Spirit to be a good brand of women's shoes. Wear your whole interview outfit for the day of a practice talk. If your blouse is see-through, you'll want to find that out before the real interview. Remember that new invitations might come in while you're on the road. I had only packed one outfit suitable to wear in warm locales but got invited at the last minute for a two-day interview in Los Angeles and had to improvise.

You will also need luggage, preferably a garment bag if you're bringing a suit. If you can't borrow one, use an opaque zip-up vinyl bag, which you can buy where you bought your suit. I bought some luggage with wheels and found that very helpful, although you shouldn't fool yourself into thinking you'll never need to carry it. Don't buy luggage at a posh store. You can get it much cheaper at a discount store, such as Target, or through AAA. I got a three-piece set of Samsonite luggage for $129 and brought the tote bag and the wheeled carry-on with me. While I sometimes checked the "carry-on", particularly when it was overstuffed and couldn't fit in the overhead bin, I was glad not to have a full-sized suitcase to lug. You often have to walk quite a ways from the gate or baggage area to the rental car. If you'll be making a number of trips, make life easier for yourself by keeping a toiletry bag stocked. If you're not sure whether to pack something, ask yourself how much it weighs and how inconvenient it would be not to have it. For example, bring a nail clipper, earplugs, and extra pantyhose. Don't necessarily bring enough reading material to last you the whole circuit, because it's easy and inexpensive to buy books and magazines as you travel. For more information on traveling light, see the Travelite FAQ. I don't recommend traveling with a bare minimum, however: Bring things that will make you feel comfortable in unfamiliar surroundings, such as a copy of your favorite tape (to play in rental cars) or a stuffed animal. Consider packing an empty duffle bag, in case you accumulate more stuff as you travel, as I did.

Be sure to think about your health needs when you pack. I was glad that I brought Pepto-Bismol, which was good for the nervous stomach I had before my first talk, and Source Naturals Wellness Formula, an echinacea-based tablet that I took when I felt particularly run-down to boost my immune system (even if only by the placebo effect). If there are any infections that you are particularly prone to, you might want to bring treatment. In any case, bring information on how to contact your doctor. If you catch a cold, medicine won't help (except for the symptoms), but if your mucous turns green, you can call your doctor for antibiotics, which will probably help.

You'll need accessories to go along with your spiffy dress clothes. All the briefcases I saw were very expensive (hundreds of dollars), but I found a nice-looking fake leather portfolio for about $15. I don't know if you can get away with a canvas bag from a conference. I'd advise women not to lug purses. The dress I bought was sufficiently low-cut that a scarf was a necessity, as were safety pins, which I neglected to pack. Douglas Adams claims a traveler should never be without a towel. I found a scarf even more indispensable. I draped it over my arms to prevent sunburn when taking a long walk on a Southern California campus, I knotted it over my chest when I discovered that the front of a dress was flimsy, and I used it as a belt to increase the formality of a waistless dress.

If you live alone, you'll have to make arrangements for mail that arrives when you're gone. In the month before your trip, pay twice the amount on your utility bills if you can afford to. For example, if your phone bill is $30, send in a check for $60. That way you can safely not pay the next bill. (I'm presuming that your utility bills will tend to be lower when you're away.) Telephone any mail-order clubs you're in and tell them to suspend your membership until April 15. Arrange for a neighbor to pick up your mail and forward anything important, or have it held at the post office. I notified the post office that certain friends could pick up my mail. One did after a few weeks and forwarded things that looked important. I also had a friend watching my mailbox at school for job-related material. Since you may not be able to get to your email very often, you might want to have a friend check it for you. Keith checked my email while I was traveling and faxed important things to where I was staying. Make sure your office-mates have your cell phone number, and tell them to give it to any callers. It is crucial that you remain reachable during your job search.

The most useful information about each place can usually be found on the Web. Print or download the home page and the pages of all the people you expect to meet, and read them on the airplane. If you're in touch with other job-seekers, you may be able to pick up gossip about places they've been (or share gossip about where you've been). If you have the chance, learn to pronounce any unusual names of people in the department you think you're likely to meet. You can do this by asking around (either someone who knows the person or knows the language the name comes from) or by calling the person's office in the middle of the night and listening to their voice mail message.

Other important information that you'll want to keep on hand:

The most important preparation is probably emotional. In addition to figuring out what job you really want and not just what you think will impress or please others, you also have to deal with the fear of rejection. Many of us getting our PhDs have never been significantly rejected academically or professionally. You will get rejected on your job search. Most places you apply won't invite you for an interview, and most of these places won't make you an offer. Remember, if you never hear "no," you're not aiming high enough. Don't be afraid to risk rejection, and, when it comes, don't take it personally. Something that helped me keep perspective was a 1996 rejection letter from a third-rate university to Prof. Richard Karp, which was posted in the mailroom at University of Washington. Karp has won a National Medal of Science (1996) and the ACM Turing Award (1995), the highest award in computer science. If even he gets rejected, I think I can handle some rejection too.

Interview Trips

Before the Interview

When you arrive into town, usually through an airport, you will either be picked up by a faculty member, get a rental car, or ride a taxi or shuttle to your hotel. Often, you'll have dinner with your host the night before your interview. A few places even have receptions for a candidate at a professor's house. Attendance at these is greatest early in the season. These events can be a little awkward, especially if graduate students are present, in which case there will be two circles, one of faculty and one of students, forcing you to socially choose between the two. I'd advise spending at least half the time talking with faculty and at least ten minutes with students. While talking to the students is easier than talking with faculty, remember to behave professionally. This is a good time to ask students what they like best about the department. You can ask them more delicate questions the next day when the faculty isn't in earshot. Similarly, don't ask faculty members any delicate questions at a reception.

Most schools will put you up at a big hotel, although a few use inns or bed & breakfasts. If you are traveling with your partner and are not married, you will probably be asked if you want one room or two. Resist the temptation to make a smart remark and just answer the question ("one room will be fine" or "we'd prefer two rooms"). One time, Keith and I were given the choice of a hotel or a bed & breakfast. We chose the b&b and found it very pleasant. When we came home from a tiring day of interviewing, our hostess prepared tea for us. Breakfast was less of a production than at a hotel but at least as good, and we took the opportunity to talk with our hostess and learn more about the region. The disadvantages of b&b's are that you might not have a phone in your room, you may have a shared bathroom, and there might not be high-tech amenities like fax machines, Internet access, and voice mail. Whether you're at a hotel or b&b, you can expect to be able to request an iron and ironing board.

When you check into a hotel or rent a car, ask if they grant frequent flier mileage. You may also want to ask at check-in what's included. Usually, local calls are free or paid for by your hosts. You may have to pay for long-distance calls yourself. Room service is usually covered but not always, and there may be a daily limit on what you can spend on food a day (which is only likely to be a practical issue if your partner is accompanying you). I don't know whether laundry or dry-cleaning are ever covered; I didn't use them. It's simpler to buy fresh pantyhose and to wash underwear in the sink with bar soap, letting it air dry overnight.

It's good to have lots of one-dollar bills with you at a hotel, since there are so many times tips are expected. When possible, such as with meals at their restaurant or room service, include the tip on the bill, so the school will pick it up. I never tried getting reimbursed for tips not associated with a meal or taxi bill. You should leave a tip of $1-2 per day for housecleaning, either daily (preferred) or at the end of your trip. Make sure you place the money in such a way that they know you mean to give it, such as on an unmade bed or with the free toiletries. While you can easily get away with not tipping housecleaning, remember that you're in a privileged position interviewing for these prestigious jobs and that these people have to work very hard at less interesting work for less money. Similarly, you should tip cab drivers at least 10-15%, more on inexpensive trips. If your partner is with you and you're talking about how rich the two of you will be, you have to tip more. Be sure to always request a receipt and to record the total with tip. You should also jot on it the date and some note that will help you remember what trip it was for when you write up your expenses. If the cab driver says he or she is out of receipts, make your own receipt on a piece of notebook paper and get their signature.

The Interview

You will typically be met at your hotel in the morning by a faculty member, around 8:30 if you're just getting a ride, earlier if you're having breakfast together. Your day will be filled with meetings of 30-45 minutes with professors and usually one or two meetings with a dean, provost, or president. You may or may not meet with students, possibly en masse. Of course, you'll give a talk on your research or, at an undergraduate institution, teach a class; some schools have you do both. You will have lunch and dinner with people from the department. The rest of this section discusses some general principles, such as attitude, and specific activities in more detail.


In the preparation phase, it is good to be a perfectionist. Not in the interview phase. If something goes wrong, accept it. If you're late to the university, whether through your own fault or circumstances beyond your control, stay calm. You'll get there eventually, and getting stressed out won't help. If you appear in casual clothes because your luggage has gotten lost, apologize to each person you meet with and to your audience, letting them know in one sentence what happened, then proceed confidently. What's less important than actually wearing the formal clothes is communicating that you sincerely tried to. While you can be a maverick in your research, don't flout convention unnecessarily in other areas.

You should always appear confident, making sure that you know the difference between confidence and arrogance. Even at schools famous for their arrogance, they do not like arrogant visitors. Here are some sample exchanges:

Question: "Why did you do X in such-and-such a manner?"
Bad answer: "Uh."
Confident answer: "For reasons A, B, and C... Although if I'd realized when I'd started that Y, I would have considered Z instead. I'd definitely have used Z if I had more memory."
Arrogant answer: "Any other approach would be really stupid."

Comment: "Your thesis seems like it's just a hack."
Bad answer: "Uh."
Confident answer: "I don't think it's a hack. The way I see it is blah blah blah. It is true, however, that it's less theoretical than some of the work that's been done, such as So-and-So's. Let me describe some research I'm planning to do on top of this (or that I did for my Master's thesis)."
Arrogant answer: "No, this isn't just a hack. It's the most important thesis in architecture in recent years."

Question: "Are you aware of So-and-So's related work on blah blah blah?"
Bad answer: "Oh no! Uh-oh."
Confident answer: "No I'm not. That sounds really interesting. I'll look into it. Could you give me a reference? Thanks for pointing me to it."
Arrogant answer: "I don't know the work, so it must not be relevant."

Meetings with Faculty Members

I felt that senior faculty members (those with tenure) used meetings to judge me or to sell the department, while junior faculty, who had recently been on the job market themselves, tried to put me at ease and invited me to ask them questions about the department. A question I asked all junior faculty was what hours they worked. I ruled out one research university when a professor answered, "This week I've been getting up at 6:30 [AM] and going home at 2 AM, but this is an unusually tough week. Normally, I go home at 1 AM." Some other questions I liked to ask were:

One professor at a school I visited told a story from his job search about an exchange he had with a very senior professor:

Candidate: "I've heard that junior faculty members here have no power, and the department is run in an autocratic manner."
Senior professor (angrily): "Who told you that?"

The candidate refused to answer the question, despite the professor's insistence. Later in the discussion, the professor untensed, smiled, and said, "I've figured out who must have told you that." The candidate wisely concluded that this was a department to avoid.

One interesting thing I noticed was that whenever I asked a male faculty member what percentage of the students in the department were female, his off-the-top-of-his-head estimate was always much higher than the actual number than I found out later from the person who had the statistics. I'd be curious if other people have had the same experience. I'm afraid I was a little undiplomatic on this subject. When a professor told me that they'd been trying to get more female students in the department, I had the temerity to ask what they were trying, putting him in the awkward position of admitting that they hadn't actually done anything.

Don't ask questions that you don't want the answer to. When I met with a female professor in another department of a school I visited and discussed the environment for women in the CS department, all of whose professors were male, she told me that one of the professors (not the department head) sometimes behaved offensively. I asked who, and she said she was willing to answer but that I should first be sure that I really wanted to know. Especially since it was a small department and I'd have to interact with everyone, I realized that no, I didn't want to know. I soon found out anyway from a student who told about some crude behavior in the classroom. At the end of the day, the department head asked me if I had any advice on how to make the department more hospitable to women. I don't know if it was politic, but I relayed the story to him, without using any names. He was surprised by the story and tried to get me to tell him who the professor was, which I wouldn't. He indicated that it was the sort of behavior that he didn't consider acceptable and would do something about if it was brought to his attention, which was good for me to find out.

Some people will offer you something to drink or a bathroom break. If they don't and you want something, speak up. The same goes for nonstandard requests, such as if you need a safety pin or to make a phone call.

Especially after you've been on the road for a while, it can be hard to be enthusiastic for a meeting, particularly if you know there's no way that you're going to work at this place. While the primary purpose of interviewing is matching graduates with jobs, an important secondary function is cross-pollenization. In your meetings, you can meet interesting people, find out about others' work, tell them about work they might be interested in, pick up gossip about other places you'll be visiting, and more. No meeting will be a waste of time if you make the most of it.

Now might be a good place for a digression on gossip. There's nothing wrong with gossiping; "gossip" (the noun) just means "information". You don't need to say anything unkind; in fact, you should say kind things whenever possible. My host at one university only had positive things to say about everyone, which made a good impression on me. My advice would be:

  1. Listen a lot more than you talk.
  2. When you have something good to say about someone, say it.
  3. If you have something bad to say, only do so if there's a good reason and, even then, tone it down, and don't let on if you're enjoying telling it.

An example of a compelling reason to say something negative would be if someone considering going to a school you're knowledgeable about asks your advice. Here are some examples of syntactic sugar:

You mean: "Almost everyone in Prof. V's group quit in disgust last year."
You say:
"I know that there are some students who had problems with Prof. V, but I don't know all the details."

You mean: "Prof. W is racist and sexist."
You say:
"Some people feel that Prof. W is most comfortable with white men, but I don't know if that's true."

You mean: "People expect Prof. X to be denied tenure in a few years."
You say:
"I'd be concerned about working with Prof. X, because he doesn't have tenure, and it's always something of a craps shoot at Drofnats University."

You mean: "Prof. Y appropriated one of his student's thesis work for his start-up."
You say:
"Prof. Y has a start-up, and there's been some concern about the independence of it and his research group."

You mean: "I can't think of anything good to say about Prof. Z, and none of the charges have been proven."
You say:
"I really don't know much about Prof. Z; I'm sorry."

Some of the most useful gossip I got was when people had worked somewhere I was considering. For example, I found out that one department I was visiting had denied tenure to the overwhelming majority of their faculty since its founding.

The juiciest gossip I received about people's home institutions was from students, who haven't yet learned to be discreet. One professor told me that there were two professors in the department who didn't get along but didn't say who they were; an undergraduate volunteered that information cheerfully, along with further details. I tended to get the most honest evaluations from students I struck up conversations with myself, although sometimes the department's hand-picked ones gave very useful information. One student I started talking with in a lab (a great place to hang out) at "Bayside" told me that his (former) advisor had been fired and was suing the university. A quick web search revealed that the school was under censure from the American Association of University Professors, a troubling sign.

Another very good way to learn about a school is reading current and past issues of its student newspaper. My host at Bayside deterred me from picking up the current issue when I was walking with him. I picked one up later and discovered that the front-page story was about a letter protesting the treatment of professors signed by all of the faculty members. Papers are sometimes accessible online, or you can also visit the newspaper office to read back issues or telephone to mail-order them.

To get the most out of your trip, keep a journal, preferably online, into which you transcribe and expand upon your notes from each day. As Mark Twain wrote, "A journal properly kept is worth a thousand dollars when you've got it done." Unfortunately, as Twain discusses in chapter 4 of the appropriately-titled Innocents Abroad, many journals get started with the best of intentions, and few get finished. I must confess that I am one of those delinquents. I went from writing a page about each place, which I shared with friends who asked to be kept posted on my progress, to writing nothing. Here are a few entries, containing both personal and technical information:

My first interview was at Mills College on Tuesday, Feb. 18.... I met with the two tenure-track CS faculty members, Susan Wang and Matthew Merzbacher. Susan got her SB from MIT around 1985. She's about to come up for tenure, and no problems are expected. At lunch, she asked me how I would supervise compiler/architecture research, and I said I'd get back to her. I had heard about Matt from Franklyn Turbak and Sean Sandys. Franklyn knew Matt from Wellesley and Sean from Williams, both places where Matt has taught. He's very nice and was a warm host, showing me around the campus and taking me to his duplex in Faculty Village, where he lives with his wife, who is pregnant....

[Note the degradation in notes quality at this later interview:] ....The other group I was being considered for was user interfaces. Dr. T was out of town, so I met with Dr. C and Dr. H. Both of those meetings went well, although my meeting with Dr. J, their boss, did not. It was my last meeting of the day, I was sleep-deprived, and I felt as though she wasn't treating my ideas with respect. Dr. K demoed his X system to me, and I showed him the Y system and suggested he see about getting their code. My meeting with Dr. P (AI) went poorly. Dr. F started by criticizing my research a lot, but things went better after I acknowledged its limitations and started taking notes on his leads: deductive databases, XSB at SUNY Stony Brook, U Wisc Madison, Ramakrishnan COREL?, Jeff Ulman's group at Stanford, Alon Levy. I had a pleasant meeting with Dr. F and Dr. S. They suggested I look at Datalog, which I could read about in Ulman's database book or in Vianu _Database Systems_ Abiteboul? They reminded me of the Z work by Prof. A, although they appeared not to have that high an opinion of it. The person I met with next (not clear from my notes) knows Lynn [my advisor] and suggested that users care a lot about response time; I agreed, also thinking that I should add user time to my measures. This person told me that Datalog differed from Prolog in having no functors, being bottom-up, and being decidable. Dr. M suggested that I look into the Verity more-like-this feature and showed me their experimental web site.

I wish I had a set of notes that included all the places I visited. Try to do a better job than I did.  At the very least, save the schedules of everywhere you visited so you know who you met with. Mark any changes on the written schedule throughout the day.

Meeting with Administrators

Meetings with administrators ranged from low-pressure chats to a high-pressure simultaneous grilling by a president and provost on where I was likely to get research funds. In general, you should let the administrator take the lead. Sometimes, the professor walking you to the meeting will have told you what the administrator wants to hear, such as when I was told, "He'll want to hear how much money you'll be able to raise."

While you should get guidance from your hosts, be sure to distinguish between their interests and yours. One search chair told me that the provost didn't appreciate how much a computer science professor needed to be paid and that I should tell her I'd require a lot of money. I didn't say anything on the subject. That was the search committee's job, not mine, or for after I'd gotten the offer. At another institute, after the first day of interviews, I told my host that it was a bad match and that I definitely did not want the job. He told me to meet with the president the next day anyway, not telling her of my negative decision, so as not to "hurt her feelings". I'm sure the real reason was that he didn't want the president to find out that the search committee had flown in someone so ill-suited to the position. In any case, I agreed not to tell her of my conclusion. (As it happened, a freak blizzard intervened, and the school closed down the second day.)

I liked to ask administrators whether any women who became a mother while on the tenure-track had gotten tenure. When no data was available on that, I would sometimes ask about the retention of female faculty. Of course, nobody told me that I couldn't get tenure if I had a baby. One president said, "I don't see why that would be a problem. That's just three months out of six years." (Presumably, he would have told a prospective father that it's just fifteen minutes.) While it's good to ask administrators questions, you should take their answers with a grain of salt. For example, it's fun to ask administrators how important teaching is ("very important") and to report the answer back to professors, giving them a good laugh.

My most interesting discussion with a provost was at a school that I'll refer to at Oswego. Because I had decided from that day's prior meetings that I didn't want to work there, I talked turkey with him, and he reciprocated. Specifically, I told him why the position was less attractive than other ones I'd seen, that there were a lot of positions open that year, and what he could do to make the offer more attractive. He listened appreciatively to what I said, concurring and volunteering further information. I asked him how Oswego compared to nearby Bayside (where I had been told that the two schools were equal). He gave a very interesting and honest answer: "Let me put it this way: Their professors graduated from here, while our professors graduated from better schools."

Giving Your Talk

There are plenty of other places where you can get advice on job talks. I'll just repeat a few tips:

If you interview at both research universities and primarily-undergraduate institutions, you may have to develop multiple talks, or at least adjust them for different audiences. When I was invited to interview at Mills, I was asked to guest lecture a class. Since I wanted to visit Mills before my imminent East Coast tour and didn't have time to prepare a new lecture, I asked if I could give a talk I had ready that was appropriate to undergraduates and said that, if I passed that first round, I'd come back and lecture a course later, without their having to pay any additional expenses (mostly picked up by other Bay Area employers). They agreed.

When I came back, I taught a computer architecture class attended by six students (Mills has small classes) and most of the department's faculty members (about the same number). I knew from the start that my blackboard technique was atrocious. (I had no classroom experience.) A student challenged a statement I made and I wasn't able to come up with an answer for what seemed like tens of seconds. I felt the class was going so badly that I was seriously tempted to stop and give up in the middle. I persevered and, to my surprise, was told after the class that I'd done well and later received a job offer. I think what saved me was having done the following things right:

  1. I had been told ahead of time that the class would be small, so I had prepared a worksheet and called on the students round-robin fashion to answer questions. It turns out the other contenders just lectured to the six students for 75 minutes.
  2. When I didn't immediately know the answer to that question, I admitted it instead of trying to bluff.
  3. I had already told the faculty that I knew my classroom skills were weak, due to inexperience, and that I looked forward to improving them. In response, I had been asked if I'd be willing to team-teach a course with a more experienced professor, and I answered (honestly) that I thought that would be great.


Meal times give you a good opportunity to see how people in the department interact with each other. I was turned off when faculty members casually insulted each other over lunch at one school. At another school, a male professor member kept touching the female faculty members more than I'd want to be touched by any colleague but Keith.

I would often be asked by the host what type of food I liked (in order to help choose a restaurant). I either named a local specialty, asked for some options, or left the decision entirely to my host. While I think a vegetarian could be accommodated at the last minute, someone who observes a vegan or kosher diet might want to give advance warning.

At lunch, I avoided heavy foods, to keep from getting sleepy, often choosing sandwiches. If others had dessert or coffee, I requested fruit or herbal tea so as to have something. At dinner, I went all out, as did the other attendees, who see these meals as a junket. The host usually asked me before ordering if I would like wine. I always said "no", since I don't drink, and added, "but don't let me stop you". This is also how I would recommend turning down dessert (not an issue for me).

The bad part about these dinners it that I was always exhausted before they were over. As socializing continued after dessert and coffee, I sometimes resorted to hints, such as visibly yawning or asking what time my first meeting the next day was, although these had little effect.


Occasionally, you'll meet up with someone who doesn't know how to behave professionally toward women. One candidate was at a dinner where an older professor cheerfully praised her looks and said that, consequently, he was for hiring her. While some people would advocate telling the professor off, I'd recommend ignoring the remark, accepting it in the spirit it was attended, or saying something like: "If you think I look good now, wait until you see my thesis results." I know it's lame, but at least it calls attention back to what is important. For guidance in this sort of situation and elsewhere, I highly recommend Ms. Mentor's Impeccable Advice for Women in Academia, by Emily Toth.

I'll take this opportunity for a digression on etiquette. The main rule of interaction is deceptively simple: A lower-status person defers to a higher-status person. What's tricky is how status is determined and what sort of deference is expected. Here are the basic rules for determining status:

Some ways that someone defers are: Note that gender is not supposed to matter in professional situations. At a crowded talk, it would be polite (but not required) for a female graduate student to offer her chair to a male professor, as would be true if the sexes were reversed. Years ago, I ran into Jaron Lanier on an elevator on my way to hear him speak at the Media Lab. Even though he's a big guy, and I was a petite young woman, he accepted my offer to "help" him with his bags (showing the same graciousness as the police officer Robert Fulghum wrote about, who, when arresting topless protesters, included a flat-chested woman).

[Later, as a professor, I was invited by the Provost to a small reception in honor of a new assistant professor.  Wanting some water, I picked up the pitcher and looked around to see to whom I might offer some before pouring my own.  Unfortunately, the two closest people were the Provost (highest rank) and the new assistant professor (guest of honor).  I didn't know who to serve first and just stood there stupidly.  Really, I should have just offered it to either or both of them.]

I think etiquette is most interesting when different people's protocols conflict. For example, after finishing an elevator ride with a senior professor at my home institution, I automatically waited for him to disembark first. Since he was raised with the rule that a gentleman defers to a lady, even if he is a Turing Award winner (which he was) and she a mere graduate student, he likewise paused for me to get off first. Did I insist on waiting him out or tell him off for being sexist (considering my being female more significant than my being a grad student)? Of course not. I got off the elevator and gave him a smile and nod of thanks. Moral: Don't take offense when someone has the best of intentions. In other words, respect diversity. This goes for inappropriate but well-intentioned compliments too.

Anyway, as a guest, you will be treated as having higher status than people in the host department, whether you're male or female. Professors will offer to carry your bags, and you should feel free to accept, especially if you have multiple or heavy bags. Don't be confused if few people offer to shake hands. If they're following standard etiquette, they're waiting for you to initiate the handshake. The only time on the interview circuit that I didn't offer my hand was when I was introduced to a man who was wearing a yarmulke (head-covering worn by Orthodox Jewish men). Because many such men don't like to touch strange women, I didn't want to put him in the situation of refusing a handshake or doing something he was uncomfortable with. I discussed this with him later (being Jewish myself), and he said that while he does shake hands with women, he doesn't like to be behind closed doors with a woman (besides his wife and relatives), which has caused offense to a secretary in the department. Just as we'd insist that a man respect a woman's sexual modesties, we should respect men's.

There are people who will tell you that it's illegal to be asked personal questions, so it won't happen to you. These people are ridiculously naive. As I suspect any female job-seeker (and some males) can tell you, they are asked. It's also not true that asking such questions is illegal, in the sense that a court would impose a punishment. What is illegal is discriminating on the basis of whether someone is a parent, for example, whether the fact is determined directly or indirectly. If a potential employer asks a candidate whether she has children, she answers yes, and she does not get a job offer, she would have to prove that it was her parental status that caused them not to make an offer in order to have any legal rights. Employers (nominally) avoid asking such questions, because their being ignorant makes it trivial to argue that they didn't discriminate.

I was asked some inappropriate personal questions, sometimes by top administrators who should have known better, such as whether I had children. Instead of refusing to answer or making a smart remark ("not that I know of"), I just answered honestly. (Asking whether I had children was always a prelude to telling me what a good place the area is for families.) I was surprised when a department head asked if my fiancé would choose a job first and then I would choose a job near his. My answer: a simple "no". You may want to think ahead of time about how you would answer likely personal questions.

After Your Visits

Submitting receipts

With each set of receipts, send a cover letter listing and totaling the charges. Some employers insist on your social security number, so you may want to include that just in case. Save copies of the receipts and cover letter. One prospective employer did need to be prodded to pay up. If you haven't heard anything a month after submitting receipts, send a friendly email asking if they need any further information in order to reimburse you.


My first offer was from Mills, which gave me a two-week response time. Because it is in the San Francisco Bay Area, where many of Keith's top choices were, we were optimistic that we could find a two-body solution, although he had not yet interviewed in the Bay Area. I explained the situation to Mills and expressed my strong interest in the position, beginning to negotiate a contract. They had offered to let me start in January and to give me a reduced teaching load my first semester. The ordinary load is 5 courses per year (two semesters), so I would have "only" had to teach 2 courses my first semester. A friend at another school pointed out to me that the benefit was less than the usual one-year reduction, so I requested and was granted a reduced load for my first three semesters (2 courses per semester). Because I was starting mid-year, the original contract given to me by Mills was ambiguous about when I would come up for tenure. I asked for and received a statement that it would be my choice whether to be reviewed for tenure after five-and-a-half or six-and-a-half years. I asked for more money in salary (even though the offer was higher than I expected) but did not receive it. [I did get a raise not long after arriving at Mills when most faculty salaries were adjusted to market rates.  I like to think I was included because I raised the issue.]  I discussed equipment needs, both technical and ergonomic, and was assured they would be filled, although no dollar amount was specified in my contract. Because of the commercial possibilities for my research (information retrieval from the Internet), I asked for and, to my surprise, received a written statement that I would retain all rights to intellectual property I create, including copyright and patents, even if I created the work on Mills equipment.

Keith and I continued interviewing as scheduled, letting potential employers know about the Mills deadline and that we needed an offer as soon as possible. One Bay Area research lab made me an offer within a day or two of my interview. I also contacted an attractive college and research lab I had interviewed at earlier to ask for an expedited decision. Both told me they would not be making me offers, which was disappointing but not devastating. One of the places had offered their faculty position to someone I too considered more qualified, and the research lab did not hire anyone that year (although I don't think they thought me stellar in any case).  [I'll admit to some schadenfreude when the professor who got the faculty position I applied for up and left the institution, presumably making them wish they'd chosen someone less flighty.]

By the end of my two-week grace time from Mills, Keith had an oral but not written offer from one of his top-choice employers, NASA Ames, in Mountain View. Since Keith and I were both in the Bay Area (him for interviews, me for a conference), we visited Mills, in order to strongly convey my interest even though I wasn't ready to sign. This also gave me a chance to see my proposed co-workers again and for Keith to meet them. I was granted another week, during which time Keith received a written offer from another attractive Bay Area company, and I orally accepted the Mills offer and withdrew my other applications. The paperwork for both NASA and Mills took a few weeks to complete.

To recap the statistics, my 100+ applications netted me about 25 invitations to interview, of which I accepted about half. As discussed above, I received two offers and two rejections before withdrawing my applications. Keith received three offers, all from research labs, before withdrawing his applications. We're glad the job search is over and feel very fortunate to have found a two-body solution.

Recommended books for job-seeking PhDs

Paperback novels about academic life to read on airplanes

Three must-reads on weddings and marriage


Mills was a good choice for me and not just geographically.  I taught small classes of earnest students and worked among a dedicated and talented faculty.  Teaching turned out to be harder than I had realized it would be.  My first semester, in which I taught two courses, I worked harder than ever before in my life, including as an undergraduate at MIT.  Planning a 200-guest wedding, which I did concurrently, was nearly effortless by comparison.  My second semester was easier, since I got a course release for participating in an educational seminar, and a senior professor co-taught my one course with me to help me improve my teaching.  Also in my second semester, I was awarded a NSF Career grant, which provided funds for research and allowed me to buy down my teaching load so I never had to teach more than two courses per semester, which I am not sure I could do well.  Later, the Provost agreed to give mathematics and computer science faculty teaching credit for the previously uncredited labs and recitation sections, reducing all of our loads to two courses per semester.  I earned tenure in five-and-a-half years, in spring 2003.  I spent my sabbatical year (calendar 2004) at Google, after which I returned to Mills with a better understanding of the skills needed for industry, and kept working at Google part-time.  After a stressful period serving as department chair 2005-2007 and during a computer science enrollment downturn, I took a two-year leave from Mills to work full-time at Google, where I am now (February 2009).

Looking back at my essay, I am amazed how many positions were available when I did my search.  I could not have gone on the market at a better time: during the dot-com boom, when CS departments were expanding and professors (and new PhDs) were going into industry. After the dot-com bust, CS enrollment -- and faculty hiring -- plummeted.  In 2008, most college and university endowments fell around 30%, causing some schools to cancel searches or institute hiring freezes.

I gained a different perspective on hiring participating in many search committees at Mills, mostly in mathematics or computer science.  I experienced from the other side trying to gauge candidates' level of interest in Mills, both in deciding whom to interview and to whom to make an offer.  One mathematics candidate with outstanding research experience and strong teaching experience insisted that he really wanted to be at a liberal arts college.  We offered him a faculty position over some other excellent candidates, which he accepted.  A few months later, he changed his mind and decided he really wanted to be at a research university.  We were devastated.  It was too late to get any of the other candidates, and we ended up losing that tenure line.

We received far more applicants for mathematics positions than computer science, perhaps because computer science PhDs have many attractive industry options.  On the math side, it was rare for us to invite a candidate without teaching awards to interview, while it was typical to invite computer scientists without teaching experience (as they did with me).  Some mistakes candidates made in their applications:
I will repeat my advice to customize applications for schools that you care about and to let committees know if you'll be in the geographic area so they can either formally or informally interview you when they might not have chosen to spend their limited funds flying you in sight unseen.

Good luck!