Another Friday, another leaving party.

Sometimes I have problems getting my head around the short-lived nature of the postdoctoral situation. What's our half-life at any one place – eighteen months, perhaps? Two years, tops? It seems you are only just making friends and thinking to yourself, “This person is so nice, I really should ask her down to the pub one evening,” and then – wham – you're clinking glasses in the local pub with her, all right, but the rest of the department is there too, toasting her departure – or your own.

Moving on is not so traumatic when you know that an exciting future awaits. Today's bash is definitely the culmination of one of academia's success stories. Golden Girl's stint here unfolded with the elegance and pacing of a John le Carré novel. Although not without its twists and turns, the project ultimately yielded glorious data and opened up a new line of research for an entire generation of students and postdocs. But the best news of all is that she will not be leaving the gold mine behind for her supervisor, Dr Derivative, to cannibalize – she's packing off the entire show to become a lab head at a prestigious university on the other side of the Atlantic. (Dr Derivative, poor dear, is putting a pretty brave face on it, but I notice that the pints are disappearing at an alarming rate.)

“What was her secret?” I asked my good friend Golgi Gal over coffee the following week, trying not to sound wistful. Golgi Gal is a postdoc in Dr Derivative's lab, so I was hoping to learn something useful. “Please tell us that it wasn't just luck, or what hope is there for the rest of us?”

“She just had the touch,” Golgi Gal said with a shrug. “Her Gilsons were jammy.”

“I reckon that good direction had something to do with it,” said Industry Girl wisely, taking a quick glance at her expensive new watch. Industry Girl is a mutual friend who left the department last year to become a team leader at a flourishing London biotech start-up down the road, but she still pops by occasionally on her breaks to taunt us with her tasteful accessories and well-adjusted nine-to-five demeanour.

“What do you mean?” I said.

“It doesn't matter how nice the car is,” Industry Girl said (and she ought to know, having recently bought a beautiful two-seater soft-top convertible). “If you don't steer it properly, you're going to end up wrapping it around a tree.”

Industry Girl explained how one of the biggest changes she experienced in her shift into the corporate environment was the sudden in-your-faceness of the scientific supervision. At first, she'd found it a bit suffocating. After all, her supervisor here at the department had been one of those sorts who is big on the conference circuit. I'm sure you know the type: so famous that they don't even bother to stay for the entire meeting; they just blow in and out for their keynote speech and leave the dust motes spiralling in their wake. Mentoring in this lab was largely performed sibling style, with the older postdocs and technicians coaching the newbies along, making sure that they showed up to work every day and at least didn't blow up the lab. Once every few months or so, people would give a research-in-progress talk and the Great Boss would listen to the story and offer a few bons mots before dashing off to Heathrow Airport.

Now, it's certainly true that an important part of postdoctoral training is easing into the mindset of independence. In many ways it's a wonderfully privileged developmental phase: you are free to strike out on your own, but safe in the knowledge that there is – in most labs anyway – a more experienced mentor around to keep you on track should the need arise. My own boss, Dr Keen, is a great example of how the relationship is supposed to work: he leaves me alone for the most part, but if I haven't stopped by his office in a while, he'll come 'round to my bench for a chat. Often, I find that I've been inadvertently heading down a blind alley, but with his observational perspective, the way out becomes suddenly obvious to both of us – and I'm off again on my own steam until the next hitch.

But not everyone is so lucky.

“Remember Zebrafish Guy?” I said, and the others nodded solemnly. Zebrafish Guy had worked with Industry Gal under the Great Boss, but hadn't responded well to the lack of feedback. None of the genetics worked for him, and after a year it was clear to pretty much everyone that his hypothesis just wasn't true. But the Great Boss didn't seem to care, and nobody else had the guts to give it to him straight and to try to convince him to abandon his project and start something new. He'd dropped out soon after that, and although by all accounts he is happy and doing well in the marketing department of a major scientific journal, it does always make me wonder if things might have been different if someone had actually intervened.

Industry Girl was obviously thinking along the same lines. “That could never have happened in a company,” she said. “We do have some freedom, but after a certain point, it's get results or go on to Plan B.” She went on to explain that once she'd got used to the idea, she'd started to see the benefits of the `tough love' attitude. “A lab is a team,” she said, “and we should all be striving towards the same goals in the most efficient possible manner.”

Privately, I'm not completely sure that academia could withstand the sort of absolute control freakery that goes on in corporate research. But Industry Girl did have a point. Academic freedom should not mean carrying along a failing route at all costs. Supervisors should feel able to intervene if things are getting impossible. And in the case of flounderers whose supervisors are too busy to notice or care, maybe we shouldn't be afraid to have a frank word with our peers. Perhaps if I had taken Zebrafish Guy for a coffee and a serious chat, I could have got him back on track. Maybe I could have bounced ideas around for a new direction, or even encouraged him to find a different supervisor. Even if the result of that chat was finding out that he was secretly longing to escape the bench and actually had no desire to prove his hypothesis and struggle on along the hard road to an independent lab position, it was probably the sort of thing we all should remember to attempt.

Postdocs, after all, have got to stick together.

Meanwhile, I've just had an email from Golden Girl. She's settling well into the Boston cut-and-thrust, and has just recruited two PhD students and a postdoc. And knowing her temperament, she'll carry on the tradition of good guidance that Dr Derivative always deployed on her.

But I hope she also packed her lucky Gilsons, just in case.