Are theses done in industry (French CIFRE) a disadvantage in academia?

Question

France has a mechanism, named CIFRE, by which a PhD student can be employed by a private company while working on their thesis. This is done in relation with a university or institute, which awards the degree. This is interesting for the student, as they are paid as a junior engineer and have an almost guaranteed industry position at the end.

I may have such an opportunity in the next year or two, in the field of machine learning/computer vision. I work well with my potential advisor and current team, and it's even more interesting in my case, as my past grades and school would not look very good on a PhD application.

If I choose to stay in that company at the end, I'm sure there won't be a problem. But I am not sure of what I will want a few years from now. I would like to keep the door to an academic career open, in case the thesis goes well, and I definitely want to keep good career options in the rest of Europe as well.

Problem

I have sometimes read that this kind of theses is only a good option if one wants to get better career prospects in the same company, and that CIFRE PhD students don't publish enough to be competitive in academia. I have also read that this could be compensated later by doing a good post-doc. These are just random opinions from forum posts, so I don't know what to make of them.

Part of the problem is that the company is small (though successful), the R&D team is very small (though they do interesting research), and I don't know if I would spend most of my time at the company or how many opportunities I would have of collaborating with people at the institute (a quite good one in France, known abroad).

Note that I might be able to get good recommendations, both from researchers who advised a small research project of mine, and from another in my current company. My grades also improved quite a lot near in my last semester, and I have a good justifications for the older ones. I could thus try to apply to "regular" PhD programs abroad, though I'm not sure how much these recommendations can compensate for old bad results.

Question

How should I decide if an industry thesis is the right fit for me? If someone knows about this specific kind of industry theses and their career prospects, I would like to hear about it too.

Answer 1

From personal experience, and from an academic perspective, CIFRE theses are a mixed bag. In some cases the company is essentially not involved at all in the thesis (e.g., they have no clue how to get involved, and are just doing it for tax optimization purposes, to keep a link with the university/school, or even out of courtesy to them or in exchange for other favors) and then the CIFRE thesis is basically funding for a normal PhD thesis with a company involved on paper (and then you'd better have good supervision on the academic side, like for a normal thesis). In other cases, the company sees the CIFRE mechanism as a way to get access to cheap qualified labor, and overload the student with work which is not always in their best interest (e.g., not publishable), at the expense of their doctoral work.

The expectations around the mechanism are also very different from student to student. Sometimes the student really wants to work at a company, and just wants a PhD to improve their career prospects. Sometimes the student really wants to do academic research, and this is mostly a way to secure funding (which might even be proposed by the academic advisor in they are collaborating already with the company). Sometimes there's real scientific value in a mixed project involving both academia and a company: that's the ideal case, but pretty rare. Sometimes the student is someone from industry which is already used to a high (aka, non-academic) salary and a corresponding cost of living, and the CIFRE mechanism is one way for them to do a PhD without giving up on their current salary. And of course sometimes the reason is that the student is not ready to choose between an academic or industrial career and wants to do something "in the middle" to keep as many options open as possible.

I don't know so much about this mechanism but from what I've heard, I would recommend:

• Discussing with the academic advisor and company advisor to clarify expectations: are they thinking that they would work together with you, or that you would spend your time between them, which kind of ratio, does the company have precise tasks for you that would be useful to them, which role would they take, etc.?
• Being wary about PhD topics that crucially require input from the company (e.g., if your PhD is about analyzing company data, but the company isn't really committed to working with you on this, you could be stuck)
• Probably making sure that you like what the company is doing and would like working with them, even for non-research stuff; or at least clarifying that you won't have to spend too much time on this if you don't want to.
• Choosing your academic advisor and topic carefully, unless you really plan to do most of your work at the company and don't intend to continue in academia afterwards.

Good luck!

Answer 2

I guess it would depend on your field but I don't think it hurts your chances. I know at least three people who did a CIFRE PhD and who then got a permanent faculty position a few years after their PhD (this was in applied math and physics). Given how bad the academic job market has been recently, I'd say that this is a pretty good indicator that they were competitive.

On a different note, many French companies have a bad tendency of assuming that applicants with a PhD are out of touch with reality and not fit for "actual jobs" (although this is slowly evolving), so having done your PhD in a "real company" would give an advantage should you choose to not continue in academia after your PhD. You don't have to stay in the company where you did your PhD.

How much of your time you spend at the company and how much time at the research institute will be part of your contract. Usually it's around 50-50. You will definitely work with the researchers of the research institute; that's the whole point.

There is also the obvious question of money. You can easily get a 30-40% higher salary as a CIFRE PhD student than as a non-CIFRE one. For some people it does not matter, for some people it does. I guess it's something to keep in mind, at least.

Answer 3

In addition to a3nm's otherwise complete answer, the more likely problem for an academic career after a CIFRE PhD would be a lack of teaching experience.

Academic research positions in France fall broadly into two categories, non-teaching at CNRS and research organizations (CEA, INRIA etc.) and teaching at universities/schools. If you fail to do any teaching at all during your PhD, you will have a very hard time being recruited for the latter. (I would assume your prospects for the former are not changed, and will mostly depend on your publications.)

Most academic PhD positions (in France) include some teaching for a yearly average of about half a day per week (your mileage may vary depending on the time you spend preparing classes, grading assignments etc.). CIFRE positions do not. Some companies may frown upon CIFRE students performing teaching; it may be harder to get in contact with the admin staff at the university to get assignments; but more importantly the practical details can be harder to arrange. In my case, the CIFRE company required on-premise presence from 9:30 to 16:30 (4:30 pm), and was located 45min away from the closest university where I could have taught anything; it would have been near-impossible to arrange for a teaching slot in those conditions.

If you care about keeping the teaching positions route open (I did not), you should discuss with your prospective PhD advisors about how to arrange such a teaching assignment upfront.