Resume tips and recruiting tips for tech in general are difficult because opinions vary widely among people who might be looking at your resume (screeners and hiring managers) on what’s good, and in general getting through the resume stage can be a very random process.
Data science as a field makes this harder because the asks on someone in a data science role vary a ton throughout the industry. “Data scientist” can mean your primary output is papers, not unlike those you’d write at a university, and you never interact with engineers, product managers, or much of anyone else, or it can mean that you do entirely product analytics and support engineering teams in day-to-day work, or anything in between. It can be almost entirely coding-based, or involve none whatsoever; and so on.
For this reason, I think the first, most important thing to do when applying for data science positions coming from academia is to determine what kind of data science position(s) you are applying for, and in particular ascertain what direct role (if any) your PhD is expected to play in your work on the team. You should be able to get a general sense of this from the job description, but can hunt around for other clues as well (if the team publishes any of their work online, if you know anyone who works at the company, etc.).
For the purposes of this answer I’ll use this (very imperfect) breakdown into three types of roles:
- “Research” positions, usually requiring a PhD, which mention research heavily in the job description and don’t talk about supporting product teams or working with engineering (or sales or finance or other non-research functions)
- Positions that clearly involve day-to-day product work but require a PhD (or are otherwise described in a way that targets PhDs even if a PhD is not strictly required)
- Positions that involve product work and don’t require a PhD—either it’s a role that doesn’t specify a specific minimum amount of experience beyond college, or it’s something like “3+ years in a data science role or equivalent experience”
I have experience seeking and hiring primarily for category 3, so my advice is best suited for that. For category 1, these positions may not be that different from academic positions, and a lot of what I suggest won’t apply. For category 2, some of the below may be useful, but obviously the parts speaking directly to comparing PhD and non-PhD candidates for how they’d do in the same role don’t really apply since you’ll be compared to other PhDs, although you might be compared to some individuals who are coming from a different private sector job, not straight from academia.
With all this in mind, here’s what I suggest regarding data scientist resume:
- If you’ve done private sector work, make sure that’s clear and noticeable, since prior history in the private sector means you know what you’re signing up for and I’ll be much less nervous about the above issues. I think technical internships during undergrad definitely are useful to indicate if you don’t have full-time private sector experience from between undergrad and grad school.
And some non-resume advice while I’m here:
- Relevant private sector internships are really really great and if you don’t have at least one and you’re not graduating right this minute, I think looking for one toward the end of your PhD can be a great move. If you are interviewing for data-related roles that normally don’t involve PhDs, from their perspective you are probably a high variance candidate: The Company is excited about the technical depth and knowledge you’d bring to the table, but you’re a big question mark as far as actual efficacy in the job. Internships are a much lower risk way for you and a company to try each other out—particularly from the company’s perspective, it’s just much easier for them to be open-minded about you. Many companies of decent size might consider internships outside of the summer, so you may not need to budget a whole additional year in grad school to do this, and if you’re lucky, you may never go through that job search at all if you are really a great fit for where you intern and you really enjoy it. (← this is what happened to me, by the way—I interned on a data team that had never hired a PhD before in that role in the spring of the 4th year of my PhD program, with no full-time prior work experience or data-related technical internships, stayed at that company full time, and my whole career stemmed from that internship.)
- When you do get the interview, think about the ‘why do you want to work here?’ questions deeply and in advance. Obviously it’s important to be authentic in how you answer these questions so you get a job you actually enjoy and will do well at, but it can be tricky to clearly state why you want to work in a place in a way that resonates. The excitement about working on the specific data set a company has is a very legitimate reason to work at a company, for example, but if you sound like you still want to write papers and just want to work at the company because it has better data than academia, that can be a sign you’re not really thinking about private sector work in the way that’s best for the team. Similarly, it’s fine/reasonable to recognize the benefits of the private sector versus academia in your answer—in fact, clearly signaling that you don’t like the pace/lack of efficiency in academia is one of the strongest positive ways you can answer this question in my mind—but you should also be ready with several good reasons why you are excited about that specific company and ideally that specific role. Leading with reasons like geography or pay difference is not a good move, even though those are legitimate reasons (and it’s fine to mention them among many). A big way to score points here (when it’s true) is if you’re interviewing for a company where you use/care about the product.
- Unfortunately, your network can matter a lot, and reaching out to people you know at a company you are thinking about is a good idea. Even if you don’t know the people super well, having any kind of referral can really help you in the process, since recruiters and interviewers know their time with you will be better spent (because via the referral they know you have some specific interest in the company rather than just dropping your resume at 1000 places, and because they think you are more likely to accept an offer if you know people at the company should they give you one, and because it’s easier to rule out some basic ways that talking to you could be a waste of time, such as your resume being a complete fabrication.) Most employees at most companies really want to refer people they know, even if they don’t know them that well—it’s fun to help people and have people you know as colleagues, and many companies offer some specific cash incentive for a successful referral as well.
- Stay positive! Job searches are high-variance, frustrating processes. Years ago I was told a (possibly apocryphal) story of a team leader at some big company in a resume review who looked at a tall stack of resumes, frowned, then threw half of them in the garbage unread, saying, “I don’t want to work with anyone who’s unlucky.” There are lots of reasons why your resume might not get noticed, and every screener and hiring manager is looking for something different. So just keep at it, and try and get data where you can—think about which roles you are getting part way through the process in, versus ones where you don’t even get in the door, as data about which types of roles to focus on, or experiment with different versions of your resume. Many companies have a policy of not telling you why you weren’t given a job, but it doesn’t hurt to ask.