Academia vs Industry: Understand the difference
Academics are increasingly searching for a path into industry. There are a number of blogs and web-platforms promising to increase employability of academics by teaching them key skills, introducing them to mentors who could guide the transition or to potential employers within the industry. At the same time, many employers consider academic talent as a junior/entry level, leading to a huge number of posts on the topic of whether PhD should be counted as a work experience.
I have done my share of moving: from academia into industry, then back to academia, then to industry and finally (but not definitively) to entrepreneurship. Both my academic experience and my industry experience span several countries and several rather distinct fields, so I developed my own vision on the subject which I think is important to share with the rising talent.
Discovery vs scalability
One of the biggest mistakes that I see people make when they talk about “academia vs industry” is the assumption that both have the same goal. It is very important to understand that academia aims at discovery, whereas industry absolutely requires scalability. So when you decide to do a PhD, you are not going to learn how to create something scalable that is aimed at helping as many people as possible, but rather you will work on something rather unique, with a great chance that the result of your work will be directly used (or read) by maybe 100 people if you are lucky (and will gather dust on the shelves of university archives if you are not).
Statistics has a perfect analogy to this key difference between academia and industry: Inference vs Prediction. Whether you are trying to understand your data, or trying to develop a predictor based on your data will drive a choice of methods and interpretation of the results. Inference is more “forgiving” in quantitative terms, but requires an understanding of the data origin and the data domain knowledge. Prediction is strict, is judged by its ability to accurately classify, and often ignores the meaning of data. Just as inference and prediction require different methods and measures of performance, so do academia and industry.
Creativity vs reliability
“Practice is when everything works, but nobody knows why; theory is when nothing works, and everyone knows why. In our lab we combine theory and practice: nothing works, and nobody knows why”
The discrepancy in the academic and industry goals creates a unique set of skills necessary for each area, contributing to PhD graduates being considered an entry-level — they often lack knowledge of established workflows or ability to comply with them, and they do not have experience working as a team member. They often forget to ask important “industry” questions: Is this useful? Can it be reasonably scaled and reliably implemented?
The “reproducibility crisis” in academia is a perfect symptom of this issue — academic path is a path of creativity, they are “painting the Mona-Lisa”, not building a machine to make 100 copies of it. Each experiment is meticulously designed to target the single question the researcher is asking, and it can take the whole duration of a PhD to answer that question (and often the answer is “No”). PhD students are taught to understand the state-of-the-art in their field, follow most recent advances, think critically and generate and test hypotheses. But they are not taught to properly document their processes, implement fail-safe mechanisms and have back-ups. And this is at the core of the reproducibility crisis: even if you asked da Vinci to repaint his masterpiece, it wouldn’t look exactly like the first one (and given the creative character, would probably look nothing like the first).
Team player?
Being part of a team is somehow completely excluded from an agenda of any post-grad academic work. Even master students have team assignments and group work. But the moment you step into a PhD — it’s pretty much just you and you alone. As a result, PhD students generally have not worked as part of the group of equal contributors for at least the duration of their studies, and have no knowledge of any project management or collaborative development tools and techniques.
Professional growth in academia
One of the key reasons academics are increasingly migrating to industry is the tremendous pay gap and difference in working conditions. It is no secret that academic jobs suck! Most academics will spend a great deal of their career on temporary contracts for 1–3 years at a time, extended as long as the law allows, for about half the salary of their industry peers, after which many will have no choice but to look for a job elsewhere. And due to the absence of “industry” experience, they often have to start at the entry level. But an academic career can be rewarding in a different way, though it requires several considerations.
Whether you are choosing a PhD or a Post-doc position, where you do it and who is your mentor will be decisive for your future. I had my share of good and bad: I quit one PhD program and completed another with distinction, I’ve been fired and I’ve been fast-tracked to promotions. So here is my most important advice to those who are considering to do a PhD and/or pursue academic path:
Choose your mentor very carefully!!!!
Since my second attempt at an academic career, I have been blessed with great mentors. They didn’t just lead the project, they embraced their role as mentors for their students, post-docs and technicians, with genuine investment into helping us to get to where we wanted to be. They also taught me a lesson that I think many have missed: The success of your subordinates is a fundament for your own success, especially in an academic environment. Unfortunately, there are not so many PIs who have learned that lesson, and PhDs and post-docs often find themselves in a position of a “cheap labor” without any prospects of professional development.
Choose your institute!
I once interviewed for a place where someone whispered in my ear: “Don’t come here! It’s a dead place!” I wish more people were so upfront about their working conditions.
Even if you love your area of research, it’s hard to stay motivated when no one else is. Academic path has no reward system, so the only reward you can find is in your environment. Top world universities and research institutes are good “go-to” places for great research environments. But it doesn’t have to be a big name to be a great place for research. Choose groups where people show a lot of enthusiasm for their research, where excited conversations about the recent paper happen in the hallway, and where extracurricular institute-wide activities are the norm.
Another hint I use to judge the place is the working conditions. Note the working stations, the equipment: is it modern? Is it renovated? Are there sleeping bags on the floor? ( last one is a joke :D but if you do see one, RUN). It can tell you whether the institute is struggling with money (and whether you should plan for some back pain because of old chairs :)) ). It might not be a definitive indicator, but you have to ask yourself — will I be excited (or even just comfortable) to spend here 10 hours a day every day for a lower pay than my peers?
Aim at management positions
And finally, if you decide to switch from academia to industry, realize that academic research provides you skills for strategic thinking, not for execution. If you want to transition to industry, aim at more strategic positions, like project management or consultant, and try to gain skills correspondingly. I see many PhDs who are trying to learn coding, statistics, machine learning methods implementation, positioning themselves for entry-level “data scientists” roles. Instead, they should leverage their PhD as an evidence of domain expertise, and complement it with soft skills necessary to guide the team to success.
