I want you so ba-a-a-a-a-ad. I want you-u-u-u. Okay, not you, really. Someone. I have a job opening to fill and I've got to find the right someone, and I can't take forever to do it, and I've got this song stuck in my head. Where's my iPod?

Recruitment. You need a job. I need someone with your skills and experience. Badda bing, badda boom, easy peasy, right? Wrong. Recruitment is tricky, sticky, sometimes icky and always picky. So much rides on getting it right, and so many things work to make this one of the hardest things we do.

For you students and post-docs, I'm not really talking to you – but stick around anyway, because you're going to need to know about this someday. And this is yet another thing that we're never taught to do, despite how vitally important it is – right up there with budgeting (which is actually easier) and organization (which someday I'll get a clue about that will actually work). In fact, knowing how recruitment works, when it does work, will also help you if you want to be recruited, so yeah, I guess I'm also talking to you after all.

The following applies whether you're looking for someone to put caps on tubes, a new trainee, the faculty member to replace the one who just left, or the Chief Scientific Officer for your new start-up; or someone to be all of the above. I don't know whether it applies to people who aren't related to the biomedical sciences, but it probably does. And this is really important, because it leads to this disclaimer – I don't know what I'm doing because, like you, I have no training in this whatsoever.

Maybe that's a good thing. Recently, I sought expert advice in the area of recruitment from a professional recruitment person, and spent a few hours listening, role playing, reviewing and filling out forms, and learning about the learned work of the experts who have far more experience, apparently, than I have. (I admit, I did wonder how they were recruited, given the results.) As it was, I came away thinking that I would have been better off spending the time watching `The Dirty Dozen'. Great movie.

Really. In this movie, one approach to recruitment is employed. Lee Marvin is an army captain faced with an impossible but important WWII heroic task, so he recruits a set of prison convicts to do the job, offering reprieves in return. This is the `anyone-can-do-it-if-there-is-enough-incentive' approach. But one of them (the Telly Savalas character) is so completely crazy that he nearly botches the plan, and therein lies the lesson. One bad egg can ruin the dozen. Some recruits are just poison.

This is a bad thing. There is a corollary of Murphy's Law (another is that all wisdom is a corollary of Murphy's Law) that says, “if you add a teaspoon of wine to a barrel of sewage you have a barrel of sewage; if you add a teaspoon of sewage to a barrel of wine you have a barrel of sewage”. It applies not only to wine, but to organizations, like yours, as well. There is no tried and true way to avoid bringing in such individuals, but here's a useful hint: if a potential recruit tells you why his/her previous position was just awful, it is only a matter of time before the recruit will be saying the same about you. We'll come back to this. Maybe the Dirty Dozen approach is not what we want.

So instead I should have watched an old episode of `Mission Impossible' (the TV show, of course, not the movies).

In `Mission Impossible', Jim Phelps (Peter Graves) would get his mission and then sit down with a folder and select his recruits, based on expertise, prior experience and, apparently, studio contracts. So he would always choose Martin Landau (and after his contract expired, Mr Spock), Barbara Bain, Peter Lupus, Greg Morris and somebody else, and they would manage to get the job done (to Lalo Schifrin's theme music). Ba ba bada Ba ba bada Da.

Unfortunately, nobody told us how Peter Graves got these crack workers, how much they were paid, or what the dental plan was like. So how do we assemble the team we want? Mission Not-So-Easy.

Here's how. The first thing is to know what you want. I do not mean the `job description', which is something Human Resources folks need for legal purposes and to make more work for me. What I mean is, do you want someone who will do what they're told, or who will perform independently and creatively with little oversight? Do you want someone with specific technical skills? Or someone with a record of productivity? Do you need a specialist in a particular field, or someone who can teach anatomy (do they still teach anatomy?)? This is crucial, and there are some points to consider in this regard.

A point: for any long-term position, go for the best possible person in terms of a) potential, b) interactivity and c) congeniality. A loose match to your needs is all you have to shoot for – get the best person. Anybody really good will still be doing great things long after their hot-hot area is cold-cold. Judge their potential on what they've done and plan to do (and how clever they seem to be); judge their interactivity on their record of collaboration (as in publication with others); and judge their congeniality in social settings (by actually socializing with them). Of course, with faculty-type recruitment, you'll have plenty of opportunity for all of these and more. Use them wisely.

I know. “Mole,” you say, “what if we really need a proteomic maestro, or someone who does really cool intravital imaging, or works with little RNAs?” Okay, if you are recruiting a facility core director, they have to be good at what you need. But faculty? The first time I was involved in a recruitment effort, some of our faculty were keen on getting someone who could generate lymphocyte lines. They got one, and as far as I know, he is still doing that, despite it not being such a big deal any more. (Oh, and my involvement? Well, they hired him instead of moi. So it goes.)

What about other types of recruitment? If possible, try to get any potential recruit together with the group – in any way you can; a visit perhaps, or at a meeting. Failing that, talk to anyone they have worked with before. But beware – some folks give very good recommendations to those who they want to rid themselves of. So how can we find out about how this person will fit? True, he/she might seem to be a Martin Landau, but is really a Telly Savalas. The latter (character, not the actor, who was probably great to work with) could cause fatal problems for your program. So here's the very least you can do – talk to them. I'm often astonished that some labs recruit workers without ever talking to them, even on the phone. But then, some labs can work with the Dirty Dozen as easily was with the MI team. They might even prefer it.

Yes, it's tricky. I recruited a lab worker who sat stone-still in my office for an hour, unable to talk, but I hired this person because when the lab came back from dinner (sans moi) they told me the recruit was terrific. I did not recruit an accomplished young scientist whose degree advisor told me could be somewhat difficult. And I never regretted that I hired a lab member who was not especially accomplished or liked by the lab (at that point), but because the recruit asked great questions and struck me as a fun person (that person is now a very successful independent scientist, a valued collaborator, and a good friend).

When it comes to recruitment, at any level, ask these questions: can this person adapt to the changing world of science? Can this individual interact with others to obtain reagents and technologies to address new problems? And will this person be a colleague — interesting and fun to spend a lot of your time with? Because that's the real deal. Nobody wants Dr Grumpy in the next office, no matter how good they seem to be; and Dr Wonderful may be great fun but only when things go well.

I heard a maxim recently that seems to ring true, and if it is true (I'm not entirely sure) then, if you want to recruit successfully – for any position – you may need to be courageous. It is this: first-rate people want to surround themselves with first-rate people; second-rate individuals prefer to surround themselves with third raters. Great candidates can frighten the faint at heart.

But how can you get the best person? Know your strengths and those of your colleagues and institution. Beat the bushes – contact everyone you know and let them know why this is a real opportunity. And when you find a likely candidate, remember that, however hard you are looking at them, they (if they are any good) are also looking at you. I have a friend, Professor Aardvark, who always sets talks for any recruit, testing them on the spot. Seems like a good idea, except that she has trouble actually getting the folks she finds suitable. I wonder why.

So, you actually have two tasks: find the right person; then convince them that this is the right opportunity. Listen when they talk and ask questions, not only to find out about them but to let them find out about you. Let them interact with the group. It helps to remember their name.

Does any of this work? All I can say is that my lab, my department and my institution are staffed with people I love to work with, and we're all doing very well. I like to go to work, which is a good thing, because I do it a lot. And I count myself very lucky that my institution recruited me.

The one you want is out there. Find them, and bring them home.