How difficult is it to make the transition from industry to tenured professor?

I agree with the other answers.  I'll just add a few observations from my own experience, transitioning AT&T Labs--Research, where I worked from my PhD in 1996 to when I joined Princeton in 2005.  First, if you go to an industrial lab, you cannot fully control whether you can make the transition back.  The academic job market is somewhat capricious, with the number of openings, the nature of your competition, and the main research areas of emphasis shifting over time.  Plus, the longer you are out of school, the more likely you are to develop more geographic constraints (e.g., because of a working spouse, or ailing parents) that limit the set of schools you can consider.  Plus, industry labs are subject to the whims of the parent company, so the nature of your position -- or even the continued existence of your position -- is not fully under your control. Second, if you are fortunate enough to be in an industrial lab that values research publications and external visibility, you can still take steps to maximize your chances of a successful move to academia.  In particular, it is valuable to (i) work with students in summer internships, and longer-term collaborations, (ii) take "unfair advantage" of your lab's resources -- whether it is measurement data or large-scale experimental infrastructure -- to do a kind of research your academic peers typically cannot, and (iii) serve on program committees, attend conferences, give talks at universities, and so on to maintain external visibility.  In short, be strategic. Third, to the extent you can pick your own research topics in an industrial lab, try to work on research problems that "walk the line" between external research interest and internal industry relevance.  Otherwise you will find you are doing two jobs at once, and that's never a recipe for doing either job to the best of your ability.  Plus, in industrial labs, resources like summer internship slots and travel budget are typically allocated more heavily to people who are making strong internal contributions.  And, the top students may be most interested in interning with you if they can have a research experience that is different than those available in academia (e.g., access to data, a chance to do tech transfer, etc.).  So, all the more reason to look for ways to walk the line. Fourth, have at least one major research project where you are the clear leader of the project, and make sure that project is larger in scope than a single publication.  Much of industry research work is collaborative, making it difficult to tell who is making the main contributions to a paper or software system.  That said, if you write multiple papers, and there is a clear "taste" or "theme" that clearly stems from you, that can shine through.  But, if you move from group project to group project, and play a secondary role on many things (whether an internal project or an external collaboration), that might not be the case.  Then, even if you have a strong publication record, it can be hard to stand out. Five, the timing of the transition matters.  Transitioning within a few years is easier than waiting longer.  Typically, you want to stay in industry long enough to have some new results to show, but not so long that you are past the normal post-PhD time for tenure.  (This makes it easier for the school to hire you as an assistant professor, perhaps with a short tenure clock.) That said, sometimes a research topic grows in importance over time, and more faculty openings arise in an area later in your industry career.  I was lucky in this respect, where computer networking (and the Internet in particular) grew in importance during the time I was at AT&T, and the specific topics I was pursuing (Internet measurement and network management) were of continued -- and perhaps even growing -- academic interest during the time I was at AT&T. My two cents is that going to an industrial lab is always risky, if your ultimate goal is an academic position.  Going to an industrial lab is a decision best made because you like that kind of research -- working closely with your peers on important practical problems with unique access to data, systems, and people that help in tech-transfer.  This is a very rewarding professional path, and can lead to many great things besides an academic job.  If you play your cards right, *and* you get lucky with the trends in academia, you *might* be able to make a jump to academia, but you might not.  You have to be okay with that outcome when you make the decision. Plus, is it worth noting that it is rarely true that all academic positions trump all industry positions.  Most people have a range of attributes they want in a job -- working on interesting technical problems with smart people, living in some range of geographies, having a certain amount of freedom, not minding whatever ways you have to sing for your supper, and so on.  Some academic jobs may trump some industry jobs, but the reverse is true, too.  Ultimately, each person puts a different set of weights on each of the factors, and these weights may change over time, too.  It's hardly a black-and-white issue.

There are really good answers here (which I've upvoted), but I want to call out one particular comment in the many that made that to me has been the key issue that people have with this transition. You can have the papers, you can be a good speaker and make a convincing case you'll teach well, you can do a variety of things…but you have to give a good hiring talk. And this is the biggest stumble for most. The problem is that they are so used to either (a) speaking primarily to management or clients or (b) speaking about their group, and each of these is a killer in its own way: when you speak only  to management or clients, you often tend to eschew niggling technical details; when you speak about a group, it becomes hard to tell what you did or can do. Neither of these is a problem in industry—indeed, both can be virtues—so you especially fail to realize that your talk is shooting you in the foot. The combination of these (and it happens often) is the worst. I know of many departments where people will shake their heads and say, "Sigh, industry talk!", and everyone knows what it means: it suffers from one or both of the above fatal flaws. Rexford's message contains excellent advice on how to avoid this happening to you. But it's important to first recognize the problem. And of course everything that leads up to this problem (e.g., if you have only been managing people without getting into technical details, you can't give a technical talk even if you now know you must). Finally, the original question asks about "industry [...] or startup". You can just about keep up a publication record in industry, especially at an industrial lab or a company that encourages research (there are very, very few of these). For various reasons, it's virtually impossible to publish research in a startup (I imagine your VCs will have some choice things to say if you spend your time on that). I know of only one person who did interesting research in a startup and parlayed that into a great job talk (William Cook, who had earned a PhD [from Brown] laying the foundations of OOP, went to industry for a decade, and is now a professor at UT Austin). He's very nearly the exception that makes the rule.

Given that most of the answers here are from people currently in academia, I thought I'd add a different perspective. First off, transition from industry to academia is not that rare as some of the answers seem to point. Following up on 's answer, even at UCSB there are other examples such as Matthew Turk, Larry Rabiner, Forrest Brewer, Larry Coldren... I could go on. That's a lot of people in a single university. If I start thinking about other universities I can easily come up with many examples such as Alex Smola going to CMU. I am not sure all the examples I cite went directly into tenured Professor, but most of them did. It would be interesting to see a more thorough analysis across different campuses though. That said, the fact that most academics have not transitioned from industry does not say much about how hard it is, but rather how little incentive successful industry people have to move to academia. I remember years ago, when I still had in mind academia, having a long conversation with someone who, after being very successful in industry had moved to academia in a top university. He gave me many reasons why that had been a bad idea and encouraged me to go to industry. I know this is not the point of the question, but I will highlight the main two: lack of resources, and politics. This gets me to the main point I want to make. The reason it might be somewhat hard to move to industry to academia has nothing to do with publications or any other form of public impact, nor with teaching experience, but rather with politics. In my experience, there is so much politics in departments that it will be really hard for a candidate to get enough support unless you have played that game before. In order for a candidate to be accepted in a department, they need to gather support from several key people. If you have a couple of people supporting you, that might get you to be considered (as points out there are likely to be hundreds of applications for any single position, knowing the right people can get you through the first cut). But, you need more than that. The only way to get this is by getting the right "recommendation letters" from people that are "well connected". This is much harder to get, although not impossible, if  you are not an insider to the system. That said, there is something departments and universities, especially in the US, like much more than politics: money. The best way you can get your chances of being accepted to increase dramatically is if you can show that you can bring money into campus. The traditional way to do this is through grants, but if you are in industry and show you can use your connections to bring money in you'll get the red carpet. The question is, if you are being successful in industry and have those right connections, would you want really want to go back to academia unless it is for some form of "golden retirement"? As much as I love teaching, I'd much rather do this as adjunct part-time professor while working in industry. And, as much as I love research, I think I could be doing much more interesting and well funded research in industry than in academia. Note I am saying I love "research" not the "publish or perish" academic paradigm.

The answers so far have been thoughtful and fairly complete, but I wanted to toss in a couple of points. I teach at a university in Japan, and an unusual one at that, with very strong ties to the Internet industry. I have worked in both industry and academia, in both the U.S. and Japan (including for the CA offices of a major Finnish tech company). I assume the questioner, like apparently all of the answerers, meant "...in the United States." I have only dealt with the tenure process in Japan, where it is very different, so YMMV. To your list of "people who have done it," I think you can add Peter Shor. He went from AT&T Research to MIT, and I assume they offered him tenure when he walked in the door; he came in as full professor (http://math.mit.edu/directory/profile.php?pid=247) (at some universities, you can be a full prof without tenure; not sure about MIT). However, his work in the 1990s ignited an entire field of inquiry (quantum computing), and very, very few of us can say that. In CS/CE, industry experience is considered valuable. Many of us have spent time in industry. If you are just finishing a Ph.D., and want that check box on your c.v., consider looking for positions at industry labs that are explicitly labeled "postdoc". Those are generally limited term, include some directed mentoring, and may be funded (even government funded) as part of a specific project rather than ongoing lab efforts. Of course, at that level you would be considered for untenured, but tenure track, assistant professorships, rather than straight into tenure. The nuance of your question, though, sounds like you might be mid-career and are considering a switch. The biggest question, of course, is, "Why?" Are you just restless or unhappy, ready for some sort of change? Is your lab headed down the tubes, either due to financial problems or because good people are leaving? Do you like working with young people, and want to give back the same way your own teachers/mentors helped you? (You don't have to answer this in public, but you need an answer to it, both for yourself and one a hiring committee will like; if those two answers are different, that should be a warning flag for you.) If you are a few years into an industry research career, and are thinking, "In a few years I want to move back to academia," then the other answers here are great. I would add to them two things: 1. Beyond publications alone, a _presence_ in the research community is a win. Are you on program committees for conferences, or the steering committee for a SIG? Do you participate in organizations such as ACM, USENIX, IEEE, at the governance/hard work level? If you are near the peak of your career and no one has any idea who you are, it will be tougher, so -- get involved. 2. Teaching as an adjunct might give you a taste of what it's like from the podium side, if you haven't done it before. I can't say for sure, but can only assume that having teaching on your resume will look good to a hiring committee. My university likes industry people to come in and teach classes in their expertise -- venture capital, Internet measurement, etc. (teaching "Intro to JavaScript" is unlikely to be inspiring to you, or to move a committee, though it would be good experience anyway). Hope this helps. Good luck with whatever you decide to pursue!

While this is presented as as a personal experience, I beleive the points I make are very generalization and largely applicable to a good majority of situations one will encounter. When I graduated in 1985, there were some excellent industry labs, and it was one of those periods where getting an academic position at a decent research university was a bit easier than getting a research staff position in some of the more prestigeous industry research labs (elsewhere [1], I talk about significant decline of industry research labs, esp. those where you can do some amount of self-directed research in CS). So during the first 9 years, I worked at three industry research labs (the last one was Bellcore, a component of the famed Bell Labs when AT&T monopoly was broken up).  I will focus on those issues that aided in my transition (but all of these are generalizable): Funding: I was able to get my own external funding ("bosses" who I felt were not that innovative and would themselves not been able to procure such external research funds wanted to exercise some control-- eg, who can work on the externally funded project that I had defined/written proposal for, which made me ask this: it seems I can get funds to do my research, so why take this interference?) Research (Publications and Citations): I had published over 5 papers per year;  luckily, one was the most cited in its area (this meant, professors on recruiting committee knew my work well, and wanted to be part of the labs I would build), and on paper, I had more publications and citations than any professors (at the associate prof. level) in the departments I was considered  (i.e., you will likely be held at a higher standard than others who are not transitioning from industry to academia!) Advising/Credential as an Educator: I  was supervising a PhD student (well, a univ. prof. agreed to be an "on paper" advisor) and had worked with a number of PhD students (typically summer interns); all had coauthored papers with me (just as one would do in academia), demonstrating some ability to guide PhD students Teaching credentials/potential: extensive lists of invited talks, colloquia, and tutorials indicated my teaching potential. The question specifically asks making transition as a tenured professor. Here I had to make a compromise-- while I got an offer as an associate professor with other attractive components (good start-up pkg include the most space they had given any faculty, etc), they could not offer me tenure because the state had the rule that for a tenure, one must have taught in a university for some period of time. So I was promised we would go through that process as soon as I have done the min. required amount of teaching (and I guess did not fail as a teacher). Finally, here are a couple of practical considerations: Timing: At different times, such transitions have been harder or easier. For example, during the past economic downturns and the times when enrollment went down (e.g., 2008 - 2014), transitions were harder, in part because a fewer number of new positions were created. I see situation improving. Also, there seems to be some periods when recurring senior faculty is more in fashion (may be because the colleges have more overall funding) than others.  Another timing issue is this: moving before you beleive your seniority and accomplishment deserves a tenure at the time of offer makes it harder to move (in otherwords, moving in first 5-8 years of your career is easier, as you may likely feel an associate prof. position with a promise for quick tenure consideration is a decent option). Positioning: What is it that you are offering that they cannot get for a person coming from an academic background (as they will be deemed safer bets)? For example, the department I moved was counting on me to bring research funds (and they got it- we have 0.5 GRA a year before I joined, and 15 a year after I joined) and strenghing the PhD program (having externally visible, higly research-active faculty members with good funding is a key aspect of building a PhD program). [1]

I work in an area that has a lot of industry jobs (e.g., ETS, ACT, College Board, Mathematica Policy Research, American Institute of Research, National Institutes of Health, etc.), though not directly in computer science. Still there are a good bit of parallels. Some folks have gone from senior industry positions to academia, but my impression is that it's becoming rarer and harder to do. The analogy of taking a world class cricket player and putting him on a baseball team strikes me as a good one. The politics of academia is one issue, definitely, but in general if you've not at least been an adjunct and sat on some dissertations, you're going to be in for a world of hurt when you try to be a professor. Teaching and being a graduate mentor are not easy and it takes quite a while to develop those skills. If you haven't done these things, it's going to be quite challenging to step into a senior role in an academic department. Also, the kind of research that goes on in industry isn't the same as it is in academia, which is usually more basic and less directly connected to specific problems. Good industry folks know when to contract with academics. Of course, vice versa is doubtless true: Just jumping from academia to an industry job likely to be challenging, but ETS has been bringing folks in from academia recently to refill their rather depleted research division, so they must feel they can make things work, and I know of a number of folks who've left tenured positions for the higher salaries and no students. Interestingly enough, though, senior people at those kinds of places are often angling to get into universities, as they tire of the restrictions that industry often places on them. I guess the grass is always greener on the other side of the fence! I suspect that a clinical professorship is more likely to work. I currently work with someone who had a long career in industry and is essentially a senior clinical professor (that's not what my university calls his job, but that's what it is). He's a great colleague, but there are things he clearly doesn't quite know, though he's a fast study and his industry experience contributes a lot.

While the transition from industry to the professoriate is fairly common; I did it myself as a 40-year old, the exact scenario of the question would be exceedingly rare at most institutions. Giving tenure to someone who has not come from academia is kind of like giving a long-term major league baseball contract to someone who has been a very good cricket player. Tenure is forever and is has become much rarer during my time in academia. Currently, fewer than 30% of all the new hires are tenure-track. The number of people hired into tenure is miniscule. In fact, it is not all that uncommon to hire a tenured professor from a third tier university and require them to undergo the tenure process again at the new place although often accelerated. Universities are more likely to hire proven performers from industry into full-time non-tenured positions that function like positions of rank without actually being ranked. Many of those coming into academia from industry have very little appreciation for tenure in any case. Their career in academia will typically last four or five years. They are often more interested in teaching than in the sort of research that career academics usually do. They end up being very valuable resources to a department due to the depth of their real-world experience. I helped hire several of them.

It's very difficult.  I don't know of anyone who has done it, although I'm sure they are out there...... I do know a lot of people that have jumped back to academia after being in industry, but you usually end up being a non-tenured adjunct professor or research scientist.  This is quite common.  Getting a tenure or tenure-track position is not common because.... 1) you just didn't get into the fraternity.  There are so many people that have gone through the traditional route, that they have better qualifications than you,  There are so many people that are making themselves miserable going through the standard rat race, that if you jump the queue, you are going to make a lot of enemies.  Also, if it becomes obvious that there are alternative ways of getting a tenured position, no one is going to be interested in going through the traditional route. and 2) quite frankly why do you want tenure?  If you have been in industry, you have money, you have power, you have respect.  If they let you just teach and do research that's good enough.  What's the point in making yourself miserable playing tenure politics when you don't have to. One particular problem is one of "ego".  There are so many applicants and so few positions, that in order to get a tenure track position you have to be meek and humble, and to bow down before the committee and do exactly what the committee wants you to do.  Once you have been in industry, you tend to develop an "ego" and this will clash with the other big egos in the search committee because you are not showing the proper respect and deferrence.  When they ask you about your publications, and you respond that publications are a waste of time and most academic papers are useless, that's not the answer that will get you the job. In other contexts, I've gotten into some nasty online screaming matches with tenured faculty which came about because they didn't think I was showing them enough respect, and in fact, I wasn't because I didn't think they deserved it.  Academia is a very small community and just by doing that, I think I insured that any application that I have would be burned before being read. Getting an adjunct position is a lot easier, because you don't care about climbing the ladder, so if you have ego/personality conflicts with one school, you can go to another one.

The three answers (at the time I'm writing this answer) cover the reality quite thoroughly.  I'll add that numerous researchers with PhDs that worked at Bell Labs have made the transition, especially those with significant inventions (e.g., the transistor).  One contemporary professor of my acquaintance who made the switch is Kurt Keutzer, who was the CTO and Senior VP of Research at Synopsys (after his time at Bell Labs) and is now a professor at UC Berkeley.