I've looked at the Wikipedia profiles of a lot of highly successful academics and researchers in my field (Computer Science/Artificial Intelligence) and it seems like nearly all of them jumped into their PhD studies straight out of undergrad.

Assuming there are a significant number of people that work for a few years before doing their PhD, this gives me the impression that folks who work in industry for a few years before doing a PhD tend to be less successful. For the academics, researchers and PhD students out there, is this an accurate assessment based on your own observations?

I've obtained a job at a FAANG tech firm as a software engineer and plan to work there for two years before heading off to do my PhD in Machine Learning. What are the pros and cons of my decision?

N.B. I've attended a top-10 school (Times and QS rankings) for undergrad and have been very fortunate to meet and interact with some truly world-class researchers in my field. While I have not yet produced top-top research papers, I think I should be able to get some decent references from the professors and scientists I've interacted with at my university and at the start-up I'm interning at.

  • 1
    A potential major benefit is that you can learn and grow a lot as an engineer while you're working, and thus be more prepared to get a lot out of your PhD. – littleO Aug 11 '19 at 9:05
  • That was my impression entirely. Life expectancy for people living in the developed world is typically close to 80 years old, so I've never been in a huge rush to get a PhD straight out of undergrad. My impression was that engineering skills developed over my two years as an engineer would give me an edge of the other PhD applicants I'll be competing against - especially in my field where implementing (programming) your ideas is a big part of the research-to-publication pipeline. – UnchartedWaters Aug 11 '19 at 23:54

I worked in industry for 3.5 years before starting my PhD. I believe I've been more successful because of this experience. However, there are advantages and disadvantages.


  • You will be more mature and have a better idea of what you want in life (which may not include a PhD).
  • Better understanding of the needs of industry which can influence your research choices
  • Connections and business soft skills (being able to sell your research effectively is very important for example)
  • Often a better work ethic (ie. used to working regular hours, etc)


  • It can be difficult to go back to being a student after making a real salary
  • Rusty study/exam writing skills
  • You will be older than your colleagues
  • If your goal is to remain in academia you may end up getting on the tenure track later with potential consequences for stability, starting a family, etc.
  • If your goal is to return to industry, the 5+ years of work experience you will miss out on is often worth more financially/promotion wise than a PhD (but this can be field dependent)
  • You will need to maintain connections with suitable academic references. As time goes forward they may not remember you very well. Strong references are important for getting accepted to top schools.

Many of these things will depend on what you do while in industry and what you want to get out of doing a PhD.

My biggest piece of advice for time spent in industry: take care of the people part - make connections, work on networking, learn to be effective in meetings, learn to sell ideas/products, how to deliver for clients. Watch and learn from senior people (directors, CEOs, etc). These things will help you be successful in academia and business. Don't just sit at a desk coding.

| improve this answer | |

I think that you might be observing selection bias.

The likelihood of someone being a highly successful academic correlates with them liking to do research in that field. This means that there’s a good number of researchers who are not going to industry because research is their passion, and you tend to do well if you have a passion for something (and possess the relevant skill set).

Industry can be really tempting, so once people (talented as they may be) go there they often stay; thus, even if you have the potential to become an outstanding researcher it’s unlikely that you’ll be working at a job that fosters it. Most industry jobs do not reward you for pursuing research; even those who do tend to focus on stuff that’s relevant to the company (or keep results proprietary) which somewhat limits your capacity for a free and outstanding research career.

On a somewhat unrelated note, many researchers in ML/AI are actively (and aggressively) wooed by industry. In some cases academic institutions are unable to compete with the benefits afforded by industry, resulting in outstanding researchers leaving academia for industry.

Two years of relevant experience can be very beneficial. It gives you an eye for what problems are ‘real’, and what models are likelier to be adopted. If it’s only two years (rather than, say, 15) it will be a minor issue in terms of adapting to an academic environment again, though your salary will drop obviously... One of the key difficulties I’ve observed with students who had gone to industry is a certain aversion to theoretical research (which is avoidable if you want to focus on empirical work), and disappointment in the lack of immediate rewards (while you do get stuck while coding at times, it’s less common for you to realize that 3-6 months of work were a complete dead end). If you’re ok with these issues I think you’ll be fine!

| improve this answer | |

Not the answer you're looking for? Browse other questions tagged or ask your own question.