This post summarizes advice I’ve given several people who are looking for jobs in industry just after finishing an academic research degree. Here, I want to be helpful by not just getting you any job, but a job that is a fit for you and you excel in. It isn’t the best term, but I’m going to call fresh PhD grads looking for a job “junior researchers” for the rest of this post.
Important Context: 2010 – 2021 was an consistently good economic market for tech companies and the people working for them. Companies were much more willing to have riskier, long-term research in their portfolio, and in general to take on the financial burden of hiring full-time employees. Any advice about job-hunting you get from people who last were on the job market during that time should be viewed as over-optimistic (this includes me).
My experience includes many smaller startups, and research and product organizations at big tech companies. I’ve been in the regular hiring loop for both, as experienced interviewer responsible for assessing junior researchers. I don’t have much experience at mid-size companies, or non-for-profit research orgs.
How work is done in academia is kind of weird. You can’t know this because you’ve spent years in it, likely with no other experience before it. Compared to the real world, academic work is:
- unusually solitary
- outcomes are self-evaluated
- owned by a single person from beginning to end
- rarely cancelled or pivoted
- focused on the “new” instead of improving existing processes, or convincing a group of existing knowledge
As a researcher, you are very good at creating or synthesizing new knowledge and working in areas of uncertainty. This is why society trains researchers. However, to apply it in the real-world, you must now become self-aware of your own alienness.
At large companies, divisions are labelled as either “product” or “research”, with research sub-divided into publishing and non-publishing research.
Working in product means you’re making changes on a timeline of less than 1-2 years, with massive amounts of data and impact on real users; this can be very fun and validating, coming from academia, to feel in the thick of it. However, your day-to-day work often reactive to put out fires from crises as they come up, and there’s less of an opportunity to explore in long-term projects. Whether this works for you personally depends on what cadence of project length feels good. You could always try product (since it contrasts to what you’ve done before) and switch back to a research cadence if you don’t like it.
Research orgs work on projects that could affect the course of the company in 5-10 years, or more. If an org prioritizes publishing, it is an arm’s length from the rest of the company, and is working effectively as a PR/recruiting wing, while also keeping experts on staff so they can be easily called in to advise on problems within the company. Non-publishing research orgs aren’t necessarily secretive, they just prioritize doing more “real” work that integrates into the companies strategy.
You should have a good idea of whether publishing is important to you for your next job. Sometimes when you’re talking to the research wing of a company, they will say “you’re allowed to publish, but it’s not measured as impact on your work”. This means they want you to take the job, but won’t give you time/resources to write or publish papers, so you’ll be doing it on your own time. Getting approval to publish papers is more time-consuming at a large company than in academia, as you will have to take to IP lawyers, public relations, and any of these people, even if they don’t understand your work deeply, may block your published work or insist on major edits. So you should assume that if publishing is not considered a part of your job, then it will be too hard to do.
Small startups are not considered enough by junior researchers! The pay is going to be less, but the amount of autonomy and seniority you get are much higher. Management can be dysfunctional and inexperienced compared to professional managers at large companies, but since you’re one of the few people doing the work, you have a much greater ability to define how you spend your time. At a startup, they are so resource-strapped they care about the outcome of everyone’s work. An upside of this is that your project will get a lot of attention and real feedback, compared to a large company where you could languish working on something irrelevant that only exists as a Plan D in the company’s portfolio. Of course, a startup could completely fold by running out of money, pivot to something that has nothing to do with your expertise (I’ve had both happen), or get acquired. But if you do work for a startup, you should do it for the positive work experience and treat any stock options you get as lottery tickets that have no real value.
Large companies are much more risk-averse than startups for hiring. Since large companies have a huge amount of resources, this is often thought of as unfair or unkind. Important idea: it is not a useful framing to weigh a corporation’s activities against kindness; this is an unrelated axis. In reality, a large company’s systems and processes are so complex, it can take 3-6 months to onboard someone into them until they start doing useful work. This means it takes that long for both the company and the employee to tell whether there is a fit between them. If there isn’t, this is incredibly costly to the company, since they need to move the employee into a new role, or the employee will quit, or they will need to be laid off or fired. Startups are much more impatient, have simple systems, and so will hire you more impulsively.
I had someone come to me recently, excited that he could get a contractor role at a company he previously was trying to get hired as a full-time employee. At large companies, contractors are second-class citizens compared to full-time employees to a great degree and you should be aware of this. This is not true at startups, where contracts are usually a quick way to get someone in the door as a full-time employee later. If you’re a contractor at a large company, they view you as potentially temporary, possibly the first to be axed in case of downsizing, and this is out of control of your direct coworkers. You may not be given as senior or as long-term projects as you’d like. Let’s say you get your “foot in the door” as a contractor and feel lucky that maybe if you do good work, you’ll be upgraded to full-time. Unfortunately, the company has no real incentive to ever upgrade you to a full-time employee at this point; you are already working for them, so why would they? If your goal is to become a full-time employee at a specific company, it would almost be better to keep yourself financially afloat and pad your resume by working elsewhere, such as a startup or research contract.
As I see it, these are the 4 major activities of a researcher:
- Vision Writing/Lit Review
- Field Research
- Quantitative Analysis
For your academic work, you’ve had to do all of these at some point. In companies where there are more human resources, they expect everyone to specialize and so the distribution of work is expected to be uneven. You will not be hired if you say “I want to do a little bit of everything”.
Try, right now, to divide 100% between the 4 activities for (1) the work you’ve done so far, and (2) the work you’d like in your next role.
I see these activities as on a distinct axis from topic expertise. You might know a lot about computer vision, machine learning, or psychophysics, but your hiring committee needs to know how you want to spend your time leveraging that expertise so they can fit you into a team.
For me personally, the line between the roles called “UX Researcher” and “Research Scientist” are blurry, because I like doing interviews and fieldwork, but also technical prototyping and analysis. But most companies consider these separate roles, with UX Researchers doing primarily interviews/writing, while Research Scientists do analysis.
Whether in job interviews or in casual conversations, I find junior researchers aren’t great at describing themselves to help someone determine how they’d fit into a team.
Don’t say you’ll do anything! It’s not about avoiding appearing desperate, it’s about giving useful information to the interviewer. If you keep saying you’re willing to do any type of work, it doesn’t give the person speaking to you any more information at the end of the conversation than they had at the beginning. This can actively be annoying to the person trying to get to know you, as they may be genuinely interested but you are wasting their time with vagueness.
Don’t just describe what you’ve done, also describe what you want to do, possibly in contrast to what you’ve done.
Don’t just describe research outcomes (for you, a published paper), describe your work. Contrast two descriptions of the same project:
“I studied behaviour of prairie voles over 2 years”
“I learned how to effectively and inexpensively deploy and maintain tracking hardware that survived mud and rain, and I analyzed time series data that included long gaps where I had to build probabilistic models of behaviour during those gaps”
It is unlikely you will be hired directly based on your prairie vole expertise, but in having to solve practical problems to get actionable data on The Prairie Vole Problem, you have earned valuable experience that is transferable to other problems, such as study prototype wearables for FitBit or a similar product.
Hope this was useful, good luck with the hunting!