One grad student's approach to making this simple, but notoriously difficult choice.

Academia or Industry? A Professionally Profound Choice

This is the third in a series of posts to unwrap my thoughts on this oft-asked, simple, but notoriously difficult question. The question is framed dichotomously, but what it really is getting at is something deeper: You’ve deferred being an adult for a while now, but what do you want to do with your life?

In the first post, I briefly toured the many options, even those not directly captured in the question. In the second post, I drilled down into the two major options. The outcome of that second post was, basically, that it comes down to personal preference because the two are very different and neither is actually “better”, despite what you might otherwise hear.

But, actually, the choice is more than just personal. Your professional development will be markedly different depending on the choice you make. The choice really is one of immense professional importance, and probably should not be taken lightly. In this post, I want to discuss why.

Disclaimer: I want to just remind everyone that I speak from the perspective of a computer science Ph.D. student at a highly competitive Computer Science school who has done 5 industry research internships — 2 at Microsoft Research, 2 at Facebook and 1 at Google. So, your mileage may vary.

This Is A Professionally Profound Choice

This choice that you’re making — the one of academia or industry — will set the course for the rest of your career. So it is undoubtedly important. But there are are also some non-obvious consequences that I want to bring up.

Certainty of Path

In today’s world, if you choose academia [1], your career path is typically more “certain”. You will join a university, stay there till you get tenure, get tenure, and then remain there for most if not all of your professional life.

Okay, yes, there is some budge. As I mentioned before, you will have summers off and the occasional sabbatical. Most summers, though, you will probably stay at your university and continue doing your research. You’ll choose to do that, probably, because you will have fewer service and teaching and other miscellaneous responsibilities over the summer. So, you’ll actually be able to dig deeper into the work, yourself, instead of just delegating to your students.

Sabbaticals are a lot longer, so you typically have more options. But, if you are like most academics, you will probably spend them doing one of three things: (1) spinning off a project into a start up, (2) writing a book, or (3) temporarily joining industry. If options (1) or (3) take off, then maybe your future path will be less certain, but in most cases you take option (3) to forge stronger bonds with industry (so that you can get cool devices and data and money when you return to your full-time job), and you take option (1) just to bootstrap the process but ultimately intend to relegate the day-to-day responsibilities of the company to someone else (a non-academic co-founder, say).

Another potential inflection point in an academic career is the decision to pursue an administrative role—i.e., become a Department Head, Dean, Provost or a President. And maybe you’ll switch universities once or twice because of a spousal conflict or because the university you are at is weak in your field.

But, basically, your career progression as an academic will be fairly predictable. There may be brief periods of uncertainty every now and again, but there’s a good chance that where you go to work after you graduate is where you will be 30 to 40 years from that time — at least if everything stays as is.

If you choose industry, your career progression is less certain. Where you work after you graduate is likely not where you will work 5 years from graduation, let alone 30 or 40 years from then. Non or partial commitment is a tenet of the modern tech industry job. Even among my friends in the tech industry, I know relatively few who have remained at the same company for more than 5 years.

This is partially because a tech industry employee now has more choice than ever before in employer and role. There are hundreds of tech companies at all stages vying for engineering talent of all sorts: Research scientists, hardware engineers, software engineers, machine learning specialists, etc. And all of these companies are jumping at the opportunity to hire capable engineers for any of these roles. Accordingly, your experience at one company can offer you a negotiation advantage at getting an even better position at another company. I’m not sure how long this will last, but the need for more engineering talent is not poised to slow down for the foreseeable future [2].

In some ways, the unpredictability of a long-term industry career, especially in something as fast-growing as tech, is exciting. You are not anchored to a company, a role, or even a location unless you choose to anchor yourself. But, in other ways, the security and consistency of a long-term academic career is comforting. Academia changes very slowly. An academic job today is likely to last a lifetime, and you can basically see a shadow of your future by looking around at the professors in your own department. In academia, you don’t have to think so much about your next career move: You only need to do good work.

Choice Permanence

If you choose industry, with few exceptions, your choice is permanent. It is unlikely that you will be able to re-enter academia unless you find an industry job that lets you publish frequently. But, even if you find that job, publications are still unlikely to be your top priority. If you work at Google, for example, publications matter very little in your performance evaluations unless you are a Research Scientist. But Research Scientist positions are few and far between, and even if you get one, you will not have the chance to progress very far up the chain of command unless you switch to an engineering or product management role.

If you get a more “academic” industry position — for example, a Researcher position at Microsoft Research — then you have a fighting chance if you want to transition into academia later in your career. These “academic” industry positions prioritize capital-R Research and publication in performance evaluations. However, your throughput will be much lower than your academic colleagues at a similar level because you will typically have fewer people with which to collaborate — both established researchers and students. As a professor, it is not uncommon to publish a dozen papers a year by running a lab with a handful of successful students. As an industry researcher whose job is to publish, you would be doing very well to publish at half that rate.

On the other hand, choosing academia is “less permanent” [3] because you can later choose to switch to an industry job if you’d like. Industry is always on the lookout for promising talent, and poaching a well-respected tenure-track professor would be a big win for industry recruiters. On the other hand, many of the skills you cultivate as a professor are not those that are useful in an industry setting. For example, you will be spending a lot of time, as a professor, raising money for your lab through grant applications. That skill is mostly useless in industry, unless you choose to spin off your own company.

Another example is the difference in focus between capital-R Research and lowercase-r research. In industry, you often have very little time to produce insights that can immediately help product designers and engineers. So, you simply will not be able to construct the perfect experimental design or do an extensive review of the literature or spend two weeks on exploratory data analysis to understand your data deeply. So if your ultimate goal is industry, an academic position would be an inefficient way to cultivate the skill set required to succeed in industry.


Your professional objectives will vary if you choose an industry job or an academic job. In academia, your objective is typically the same no matter your employer or field: do good, original Research and distribute that knowledge effectively. If you transition to an administrative role, maybe that objective changes to: make it as easy as possible for everyone under you to do good, original Research and distribute that knowledge effectively. This is your day-to-day, month-to-month and year-to-year objective. Maybe the specific things you do to accomplish that objective will change, but that will always be your objective.

In an industry role, you might also have a very broad objective: to help your employer accomplish its mission statement. At Facebook, that’s to “make the world more open and connected”. A Google, that’s to “organize the world’s information and make it universally accessible and useful.” However, your specific day-to-day objectives will probably be quite different. Moreover, a lot of it will trickle down to you from upper management. Maybe that’s okay. You might be an adaptable person and prefer just to work on challenging problems, no matter the problem. But, unless you, yourself, are fairly high up the management chain, you are likely to have less control over your medium to long term objectives.

Professional Development

Your opportunities for professional development will also markedly vary depending on your choice of an academic or industry job. Apart from the clear differences in development that you can expect from cultivating different skill sets, professional development opportunities are more broad and cross-cutting in academia. It may sound odd, because if you think about being detached and cloistered, you typically think of academia. But when it comes to **opportunities** for professional growth, you are typically more insulated in industry.

First, you are heavily insulated within your organization and shaped by the opportunities they provide. This is less true in academia, because you are more likely to interact with people outside of your organization through conferences, student exchanges, sabbaticals and so on. If you don’t believe me, ask yourself this: how many times does the average IBM employee collaborate with somebody from Apple or Facebook or Google? It is far more likely that a professor at Cornell collaborates with a student from MIT, on the other hand.

Second, you are heavily insulated within your role in industry. Sure, a software engineer will have to interact with user experience researcher before a product launches — but that interaction will rarely be one in which the engineer and the researcher hope to teach other skills. It is far more likely that the exchange between software engineer and UER will be transactional: The UER tells the engineer what problems she uncovered from her research and ways to fix it, and the engineer will take that feedback into consideration in the next iteration of the product. On the other hand, interdisciplinary collaborations in academia are more likely and generally have a lasting impact on both collaborators’ skill sets.

On the flip side, large corporations typically have bigger and better resources than the average university. If a professional development opportunity aligns with an organization’s goals, then those resources can take you far even if you put in little effort to pursue novel opportunities. I have a friend at Facebook, for example, who gets to travel all over the world to research the Facebook user experience. You are less likely to have these sorts of opportunities fall onto your lap in academia.


Finally, imagine, 40 years from now, what your career might look like if you choose academia versus if you choose industry. The set of signs that indicate a successful career are typically quite different. As an academic, the sign of a successful career in its twilight stages might be a massive citation count; an impressive student list; meta-awards won (e.g., Lifetime Achievement awards, lasting impact papers, Turing awards); service responsibilities offered; keynote talks given; and, a chaired professorship.

As an industry researcher, the sign of a successful career in its twilight stages might be how far up the org chart one has climbed; how much impact one has had at shaping organizational culture, products shipped, and industry standards (e.g., creating things like Map-Reduce); and, awards of practice won.

Parting Words

Which is better? Yeah, you know I can’t answer that. What I can say is that this is a professionally important choice. This a choice that will significantly, substantially and perhaps irreversibly impact your career trajectory. So it’s both deeply personal and professionally profound. But, it’s also societally significant — and I hope to touch on that point in my next and final post in this series.

One final note: If you liked this post and would like to show your support, here are two things you can do:
– Follow me on Twitter @scyrusk; and,
– Consider signing up for my mailing list below. I won’t spam, you can unsubscribe at any time, and you’ll get some content that I don’t post on the blog.

If you do either or both of those things, you’d make me happy. Thanks!


[1] By “academia”, I mean a tenure-track professorship. As I mentioned in a previous post, though, academia could mean many things.

[2] Consider the fact that President Obama just announced a “Computer Science For All” initiative.

[3] In the context of this dichotomous choice, at least.

Share on FacebookShare on Google+Tweet about this on TwitterShare on LinkedInShare on TumblrPin on Pinterest

Leave a Reply

Your email address will not be published. Required fields are marked *