12 Things You Should Know Before You Start A PhD

So you’re going to do a PhD and be a great big cleverpants.

Good for you. Occasionally, this is a good idea.

More often, it isn’t.

But those are just the odds. I don’t know you. I don’t know where you’ll land. What your work will mean. What will be important. What you can sell. Who you’ll meet. How your work positions you for a non-academic career. “Doing a PhD” is a very broad church, sufficiently broad in fact that I can bring precisely one certain thing to the collective table:

It’s a guarantee of nothing.


All that will change with 100% certainty is how you fill out an immigration form.

Damn, it’s DOCTOR Phillips now. Hey, honey! Look what I get to do now!

And believe me, that is a very satisfying fifteen seconds.

Past that? Your PhD entitles you to very, very little. The most important question people need answered is: “Will my PhD result in me working permanently in research?” Of course, again, hard to say out of context. In specific situations, absolutely.

But overall…? No. It won’t.


Because the amount of work available is cripplingly low.

Depending on who you believe, where you are on the planet, and how permanent/tenured/serious a job you want, the estimates for ‘who manages to stay in academic research after a PhD?’ range from 20% to about 0.5%.

In many respects, this is awful. Most PhDs are being carefully trained for careers they are very unlikely to have, and this unlikelihood is of a sufficient size that it renders how good or bad your Phd IS somewhat irrelevant.

In other words, a PhD can be an interesting and stimulating experience, or a mighty but worthwhile struggle, or a Sisyphean disaster, or an utter horrorshow that tears apart lives and families, but all of these outcomes dump you in the same job market for junior researchers, and in general, that market sucks.

And the worst part is before you start a PhD, no-one tells you anything about what to expect, or what to do. They sometimes try, bless them. But in general they do a rotten job.

The problem is that people who give you advice at the start of your PhD generally have a vested interest in the outcome. Think of your adviser/supervisor/boss: you’re about to work for them and do all the jobs they don’t have time for, sometimes for free. That tends to bias them a bit.

Your boss may also occasionally be not very clever, or not very experienced, or — most commonly — massively divorced from the circumstances that you’re going to be subject to.

You need to realise this right now. When it comes to the question of “what happens next for me?” your adviser and mentors might have no idea whatsoever what they’re talking about. I know their job is supposed to be something highly cerebral and they have some impressive title like Senior Professor or Research Fellow or something else equally weighty. I know they have fashionable square glasses and a hundred papers in the Journal of Opaque Neurobollocks, or two papers in The Extremely Important Journal.

I know they look and act like they know what they’re doing.



a) at some point, they almost certainly got luckyand do not expect them to admit this — expect them to think their incandescent genius was solely and entirely responsible for all their good fortune,
b) they usually started their career in a totally different environment to the one you’re about to wade through,
c) as I said before, their incentives can be skewed because you’re about to work for them and it will cost them little/nothing, and
d) no title that can be conferred to anyone is defense against being selfish, difficult, careless or an idiot.

Yes, an idiot. Kings, Queens, Presidents, Chancellors, Colonels, CEOs, CFOs, CTOs, Viceroys, Duchesses, Deans, and Donkey Wranglers have all been idiots. The present candidates for US president (“tha leader of tha free wurrld, buddeh!”) are a gaggle of unfortunates. Don’t be thinking Assistant Professor McClever is somehow immune to the title/intelligence disparity.

And don’t be thinking you know a whole lot about the process— if you’re starting your PhD, you may never have been in the job market before, ever. Even the weird quasi-job market that being a starting PhD means you enter. So, you may be short on experience, and also short on good advice.

Certainly no-one told me much that was useful beforehand. I survived the process more by bloody-mindedness than good planning.

The following is a list of what I learned.

(I never give advice, so bear with me as a qualify everything. a) Please bear in mind I’m waving my own biases like a rally flag. b) It’s difficult to give advice that crosses between the Commonwealth, European and American systems — bear in mind also that I’m luminously Australian. c) I’m more familiar with some bits of social and biomedical science than other areas. d) I’m also going to generalise a lot, which is necessary unless you want to finish reading this today. e) I have no time to make distinctions on the research / teaching divide in training and institutes here, this is already 4K words. So, grain of salt for everything, if you please.)

12. Try before you buy

If there’s a way to work in or near the field you want to do a PhD in (after undergraduate, honours or Masters) then try it. Often it’s easy enough to get a job doing Research Assistant work or similar. You’ll find out very quickly if the lab/department environment is something you can put up with.

One of the highest compliments students have ever paid me, and I’ve heard this quite a few times, is “You made me quit.”

NOT because I beat them with a length of rusty chain and ran them out of town, but because they took a realistic look at what they’d have to study, and how they’d have to work, and how hard I doubled down on work I figured I had to do, and figured out this life wasn’t for them.

(I should add that ALL of the people who were capable of making that decision with a certain degree of aggressive maturity have thrived elsewhere. Knowing what you want has a tendency to focus the mind.)

11. Everything takes AGES. Learn to plan.

Publications take ages to be published. 
Applications for funding take ages to assess…
and then even longer to become current…
and even longer than that to actually pay you.
Fellowships always start ‘the next calendar year’.
The money is ‘coming out of next year’s budget’. 
Finding work takes ages.
“Networking” takes ages.

EVERYTHING takes ages.

The only solution, unless you actually enjoy the experience of waiting which is usually accompanied by being bored and poor, is learning to plan.

So, you’re going to graduate from your PhD in July. Well, when should you start looking for a postdoctoral position? The answer is not “August”, the answer is NOW, GODDAMN IT. NOW.

10. Don’t do it unfunded or part-time.

A lot of people decide to do a PhD in their spare time, part time, partially by correspondence, or in any number of other assorted arrangements that don’t mean going to an office in a place full of other academics, scientists, researchers, PhD students, etc.

Don’t be one of them.

Just don’t.

You’re proposing something that will take forever and will kill your motivation. You probably won’t get good supervision, because it will be in proportion to your contact hours (low) and your likelihood to continue to be an effective full-time researcher (also low).

You give your twenties enough time and they’ll become thirties. You’ll end up with kids. You’ll end up with sick parents. You’ll end up fighting the university faculty to be allowed to enroll in a sixth year (BUT THE REGULATIONS CLEARLY STATE etc. etc.)

Part-timers crash and burn.

Two things about part-timers are often true.

  1. part-timers are often doing a PhD in the first place because of some personal goal — because they think it says something about them — rather than because of an ongoing interest in the subject matter, or the pursuit of a career in research.
  2. part-timers end up in a market full of PhD graduates who are five years younger, more current and more flexible, and when this happens they suffer. And you might remember from before: the market sucks.

Listen: getting your friends to call you Doctor when you meet for brunch is a rubbish reason to bother with inviting this monster into your life and sacrificing thousands of hours to it.

If you have the pathological need to call yourself Doctor, buy the title on the internet. It’s so easy. I bought my friend Bald James the title of Pope from an online church. A doctorate is even cheaper.

And if you need an ongoing project that takes forever and makes you cry, learn crochet and make a king-size quilt.

9. A lot of researchers / academics are complete bastards.

If I reeled off every horror story I’ve heard — every incompetent halfwit supervisor, every personality disorder, every Machiavellian paper-stealing shithead, every sexual harassment lawsuit in the making, every Saturday-working robot who expects you to mirror their broken social standards, every handsy adjunct, every rusted-on faculty member who couldn’t be fired because of some infuriating technicality, everyone who can’t handle their liquor, everyone whose personal habits would make a monkey blush, every Tin Hitler fiefdom-building little scientific despot — well, this would just be twenty thousand words of stuttering, swearwords and outrage. There are academics in the world who are so bad at their jobs and such pricks it physically hurts me to think about them taking the public’s money.

A lot of them are not cleverer than ‘regular’ people, they definitely aren’t special, and they often have very little commitment to a central idea of collegiality.

I’ve met full professors who, in a rational world, would fail an interview for apprentice dog-catcher.

I’ve met fewer, but still a significant amount, who in an ideal world I would be allowed to punch very hard in the face. Idiots, sexual predators, functional imbeciles and smug, awful, petty miserable people.

The point is not that they’re stupider or meaner than the stupid mean people in any other profession, but that they have feet of clay. Just like everyone else. There is no pedestal.

A lot of people say these people rise through the system because academia / research / science rewards sociopathy. Maybe it does, but I’d be very surprised if genuine sociopaths were that widespread. Sociopaths are attracted to power, and if you have the whole job market to choose from, science isn’t exactly a pathway to much of that.

But without the need for anyone to be a sociopath, there’s a center to this problem which is just cultural and doesn’t require anyone to be committed to Broadmoor — it’s that academic researchers are often old people with job security who exclusively employ young people who just happen to be crucially reliant on them, and who have little job security.

For career advancement, for resources, for publications, everything.

No sociopathy required, just normal people with a LOT of power.

If you’re thinking “well, that system probably has the potential for abuse” then you’d be correct.

Sometimes when the ‘old people’ are men and the ‘young people’ are women, shit goes queer. I’ve read about this a lot more than I’ve seen it up close, but the reports are in — your avuncular lecturer or laboratory head or visiting dignitary is often not your best friend.

But don’t think this doesn’t go for both genders in both directions, and makes problems far more often outside of the specific context of sexual harassment than inside it.

If something doesn’t feel right: keep your eyes open, document all crazy behaviour in writing, remember your boss is not your friend, and have private meetings with the door open if your spidey sense tells you to.

8. You’re going to move. Make plans now.

I might be repeating myself, but the post-PhD job market sucks.

The only blessing is that in biomedical/social science it’s probably in better health than other research disciplines. Ask a senior scientist if they’d let their own children get into science, if it’s presently a reasonable career path.

Often their response will be “no”.

Now ask a senior sociologist…

The central reason for this (apart from the fact that funding levels historically stink everywhere) is that strategic opportunities for young researchers are as bad as they’ve ever been, and competition is fierce.

The only nice part is that jobs in your sub-sub-field may well be available somewhere, you just need to widen your search to the whole damned planet.

Make your peace with this now if you want to reliably move from a PhD into postdoctoral research or similar.

Do you have sick parents you don’t want to leave? A boy/girlfriend husband/wife with a local job? A circle of friends you can’t live without? Well, you probably can’t take them with you. I’d get comfortable with this premise as soon as possible.

If this is a problem, don’t do a PhD. There are better ways to make a living where you don’t have to pack your life into two suitcases, give your pets away and leave your home full of people who might die in your absence.

Sorry this one is grim. But that’s how it is. Here’s a fierce denouncement of where several years of this will get you.

7. Diversify.

There will be specific skills which are useful for your area of expertise.

Learn them.

There will be others which are peripherally related. Learn them too.

Then learn how they might be useful in context. Learn some basic networking skills and use them to tell other people about the skills you just learned. If there’s a conference in your local area (especially a cheap or free one), go to it. Listen to the people who have skills. After they have spoken, ask them what skills are good to have right now. Go and acquire the skills they recommend.

A PhD is a training degree. It is skill training. GET A LOT OF SKILLS.

If you need more guidance on that…

6. Read job ads now.

“Skills. OK. What skills should I get?”

Go to a job board — Times Higher Education, Seek, ResearchGate, HigherEdJobs, AcademicPositions.eu — and find an ad that you’d be interested in applying for when you’re finished. Right lab, right work group, right job titles, whatever.

Now, look at the requirements.

We are looking for a postdoc researcher who has a) a genuine interest in behavioral cardiology and interdisciplinary work b) experience in statistical analysis and analytic models using different types of data, c) excellent scientific writing skills, d) demonstrated capacity to drive first author publications, e) proficiency with R, MATLAB, SAS, STATA, SPSS or similar statistical software, and f) good ability to work within a team context.
We seek candidates with a strong background in experimental physics, chemistry or quantitative biology, with an interest in biophysics of living cells. Applicants must hold a PhD degree in a relevant field. Prior experience with superresolution and/or single-molecule imaging is an absolute requirement.

That’s what skills you should get. You have three years at absolute minimum, in some systems more like 5 to 7. It’s all doable. In fact, this is one thing which is more possible now than at any other point in time. Equipment and skills are no longer heavily guarded secrets, software is available instantly on the open market etc.

5. You don’t need one mentor as much as you need a network lots of friends and collaborators.

I’ve seen a lot of people look for a mentor. I don’t know if that’s necessary. It’s not that you don’t need mentoring, but you don’t need a single paternal figure with pince-nez who gives you sage advice like Professor Dumbledore.

(Look at me using a young person reference. How do you do, fellow kids.)

Instead, you need a network of dozens of people you can work with, who have similar interests, who you can be civil and pleasant to. People you can help, people who can help you. People who can tell you where the field is headed. People who know where the work is. People who apply for money and set the job requirements of how they spend that money. A network, essentially.

They don’t have to be your best friends, you don’t have to live in each other’s pockets, you just need to know what’s up. You need to help each other.

Meet people. Help people.

4. A PhD is free time. Try to find a well that isn’t dry.

A PhD that you fund yourself (i.e. with a normal scholarship or fellowship) is about the most freedom you’ll have in research.

What I mean by this is: the money you get in most scholarship schemes is awarded to you personally, and not to your boss or laboratory or workgroup. Your program of research is often driven by JUST YOU.

This means you can — within reason — take your money anywhere and work on anything you say you want to. Obviously this isn’t always the case, but in general you have more freedom as a student than any other point in your career.

I have seen many, many people change horses mid-stream on the focus of their project. “Oh, that’s not going to work, we didn’t know what we were doing, I’m going to do XYZ.” And off they go.”We don’t have time to even think about getting that completed, it’s way harder than we thought. I’m going to do ABC instead.”

Some bosses/laboratories/departments have enough built-in flexibility to encourage this process of reconsideration after you spend 6–12 months looking for initial results.

Even if your project IS funded by a source which demands you work on a specific project, you should have a reasonable amount of freedom to work within the framework you’re given. Mould what you’re doing to your own ends. Engage with the process. Fight for your ideas.

3. Corresponding authors correspond.

Everyone is contactable by email. Don’t ever kill your natural interest in the subjects you’re researching. Get to talking to people. I spend a chunk of most days writing to people I have never met, or getting emails from them. Often this is the best bit of the day. Be part of a community. Involve yourself with the people behind the papers. The worst thing they can do is not write back. No-one’s going to resent you being genuinely interested in their life’s work.

(And a frightening amount of people are on social media these days. It’s a good way to figure out where things are at.)



In that order.

Look at these as skills.

I cannot stress this enough.

No matter where you end up, you’re going to end up handling data.

I mainly use Matlab for analysis, but I’m trying to make the transition to Python. I’ve suffered through enough C+ to make a Mex file work. I’ve played with R, and EEGLab and BrainVision Analyzer. I’ve used SPSS and SAS and Stata and Igorr and literally every major experimental programming software that exists. And qualitative analysis (something I’ll never, ever use again), I can do that.

And probably a dozen other things that don’t come to mind.

I’ve done everything from simple restructuring of data-sets to stand-alone programs and utilities (writing actual software… not GOOD software, mind you, but just stuff which works). If you threw me into ANY lab doing electrocardiology, anything biometric, exercise physiology / sports science, or literally anything in the social sciences which isn’t fMRI, I can make your rig work.

(And if I can’t, I can figure out how to do it before you can figure out I don’t know how to do it!)

This was all picked up along the way. I have a ‘formal’ education in exactly NONE of these languages/packages/techniques/whatever. I didn’t go on a course, unless you count online resources from other universities.

Everyone knows the expression “In the land of the blind, the one-eyed man is King.” — well, in a world full of scientists STILL pottering around in Excel, anyone with basic data handling and coding skills … yeah, you get the picture. The one-eyed man/woman is flat-out necessary.

When it comes to basic technical skills, some scientists are blind, deaf AND dumb. You’d be very very surprised. People whose work is blisteringly complicated, highly technical, and very impressive… are cutting and pasting their data into spreadsheets.You can occasionally save them hours of manual work with even the simplest half a dozen lines of atrociously loop-heavy, brutally amateur code. And that’s just for starters.

The comparative advantage these skills give you is enormous. You know the best part about your atrocious code that works? That it works.

Writing is the same, only it’s even more universal. You’re going to be called to write a lot of technical literature. But after you’re done writing that, write something else. You’re learning scholarship, scholar the skin off it.

I’m lucky that I genuinely enjoy the act of writing, and started doing this years ago simply because I was compelled to. I only found out writing was a ‘transferable skill’ or whatever hideous management jargon can be applied to it much later.

(And, also much later, I found myself astonished by watching other people write a) slowly and b) badly.)

1. Choose your supervisor scrupulously.

Some supervisors have excellent and obvious track records of turning out PhDs that get jobs.

This is ALWAYS a combination of a) students get good skills and then become employable (see 2), b) the lab they start in having a good reputation (see 9) and c) the supervisor having good contacts and networks with others (see 3 and 5). These things continuously trade off each other.

Other people will have shit PhDs from shit labs where they worked with shit people through absolutely no fault of their own. This can be a viciously hard academic pedigree to overcome.

In other words, choose your PhD supervisor with scrupulous care. As ridiculous as it sounds, he or she is your first step into a phase of your career that is not, NOT, NOT, NOT a meritocracy. Most jobs aren’t advertised, they’re cobbled together from a loose network of people and affiliations based primarily on trust.

So, if you’re going to do this, here’s precisely how to do it: before starting a PhD anywhere, MEET THE PEOPLE IN THE LAB WHILE THE SUPERVISOR ISN’T THERE.

I cannot overemphasise this.

Ignore the rest of this advice — even all my screaming about coding and whatnot — and you may well muddle through.

But pick the wrong workplace, and you shoot the chances of your life staying normal right between the eyes.

With the lack of bell-towers and megaphones and wire services and carrier pigeons that the present communicative environment affords, I will simply repeat myself.


If the place sucks, 90% of these people will look you in the eye and tell you. Right to your face. “If I had my time again, I wouldn’t have started here.”

Often they will say this not TO you, but AT you.

Ask a few, and you’ll know what’s up in a flying hurry. Investigate in detail. You may not even know the right questions to ask, so ask enough people until the right questions coalesce, and then when the questions start to make sense, ask them. Find the people who just graduated from the lab, or just left. Ask them too!

Weird things lurk under the surface sometimes…

Your famous institute? It sucks, everyone hates it. It’s dysfunctional, full of infighting and intrigue, and bad administrators, and other things which are going to ruin your life.

The random place down the road you never heard of? It’s a hotbed of productive activity. It’s full of supportive, clever people working to their mutual benefit. The bosses are great. “Yeah, we got lucky” says everyone who works there.

That place with the job ad up? Depends on who you work for.

Famous super-scientist guy with a great reputation? He’s a bastard and a tyrant. His students hate him. He fudges numbers. We can’t wait to leave. Nearing the end of his career. Just sits on committees and does invited talks and wastes our time.

Unknown and unremarkable science lady you never heard of? Best boss in history. Highly competent. Understands life stress of being a grad student. Friendly. Demanding, but fair. Her students get jobs. She’s great. I even like her kids.

Not a single detail of anything like this will ever be written on any marketing material, anywhere. But finding out these details before you start is your single most important task before starting.


No good deed goes unpunished, and your decision to try to improve the world and devote your life to your innate human curiosity and the betterment of humanity is odds-on to turn out horribly. We picked a fairly miserable point in history to do this.

Good luck. It sucks, but it’s the only science we got.

For more pomp and circumstance:

My Facebook (it has writing)

My Podcast (it has science)