I’ve hired 3 cohorts of data science interns — here’s my advice on getting an offer
I’ve Hired 3 Cohorts of Data Science Interns — Here’s My Advice on Getting an Offer
Resume and interview tips for landing a data science internship
In my current role, I’ve had the responsibility of reviewing resumes, performing interviews, and making data science intern hiring decisions for the last three years. As my group is prepping for our fourth cohort of summer interns, I thought it might be helpful to publish a few pieces of advice based on my experience and observations.
This isn’t a comprehensive guide to getting a data science internship — a lot of other people have done great work on that subject. This is a hodge-podge of advice that comes from my experience looking at hundreds of resumes and performing dozens of data science internship interviews. My hope is that you can find a few unique tips that can help round out your resume and interviewing skills. I’ve put them in order of my opinion of their importance. Let’s jump in!
Tip 1: Differentiators are always important, but they are king for internships!
You always want to differentiate yourself and your resume when looking for any kind of job. This often naturally happens through work experience. But, for internships, work experience is not very common. Because of this, there is not a ton of variety in the resumes I see. Even small differentiators can have big impacts because of how similar most resumes are.
I’ll express the huge need for resume differentiation further by asking you to put yourself in my shoes at a recruiting event. Imagine you are at a booth in a recruiting event for the statistics department at The Generic University. For two hours, students wait in line to introduce themselves to you and hand you their resume. By the end of the day you are hoarse, your feet hurt, you have a blurred memory of over one-hundred faces and a stack of resumes that rivals the IRS tax code in size.
Now I have the unenviable task of reviewing said stack of resumes and deciding whom I would like to invite to an interview and whom I will pass on.
I start looking at the resumes….
Resume 1 : Jane Doe
- Master’s in statistics from Generic University
- No work experience
- Some random hobbies
Resume 2 : Joe Doe
- Master’s in statistics from Generic University
- No work experience
- Some random hobbies
Resume 3: John Doe
- Master’s in statistics from Generic University
- No work experience
- Some random hobbies
And it continues for dozens more!
You can see my problem here, right? I can’t interview every single statistics student, but so many of them look almost exactly the same! So, I scour the resumes for some kind of differentiator — anything that sets them apart from their classmates in a meaningful way. Because of how similar internship candidates can appear, even small differentiators can have huge impacts.
Okay, I think I’ve driven the point of why differentiation is especially important for internships sufficiently. Let’s get into the kind of ‘differentiators’ I look for when reviewing resumes:
- Any kind of data-related work experience (data analyst, statistics tutor or TA, etc.)
- Freelance job or two doing some kind of data work
- Some kind of a certification (e.g., AWS ML certification)
- An interesting data-centric project outside of your school work
- Writing data science articles (like in Towards Data Science!) or blogs
- Interesting data science topics for thesis or dissertations
- Development of an R or Python package
- Participation in a data centric hackathon
- Participation in a Kaggle competition
- Leadership position in a data focused club at school
Of course, this list is not comprehensive. The point is, do something impressive that other people are not doing and then advertise it!
Sometimes, candidates will think that something is a material differentiator when (in my opinion) it is not. Here are a few things that I don’t really see as differentiators:
- Projects related to coursework — I assume everyone does school projects
- GPA — Because of varying levels of grade inflation, I never really know what is an impressive GPA. If I’m in a pinch I might use it as a last-resort differentiator (e.g., two really similar candidates with a larger delta in GPAs or similar candidates from the same program), but I prefer to use the differentiators I listed above first.
- Interesting hobbies or background not related to data — I’m not saying to not put these in your resume. It can give you some personality, I just won’t give you any extra points for them.
- Being a member of a data club in school (This doesn’t tell me anything about your level of activity in a club — did you show up once for the kick off where they had free pizza or are you an enthusiastic, active member?)
- Cover letters and follow-up thank you emails — I don’t see electing to do or not do these as indicative of your potential as a data scientist
I’ve focused here on the resume because without differentiators on your resume, you are much less likely to get the interview. Of course, highlight your differentiators in the interview as well. Don’t rely on your resume to speak for itself, speak for it!
Tip 2: Express interest in data science specifically (especially in your resume)
This seems like a no-brainer, but a lot of resumes (your competition) miss the mark here! I’ll get into some ways I recommend doing this below, but I first want to emphasize who needs to pay the most attention to this section. Anyone who has a major that is not obviously data science-related (basically everyone studying anything other than statistics, data science, and machine learning) should pay special attention here! In my experience, computer science majors fall into this problem most often, so I’ll pick on them to serve as examples. They often have great resumes for software development, but don’t mention anywhere that they are interested in data science. This leaves me wondering why they applied. I don’t have time to reach out to each candidate to understand their motivation, so I have to dock points since I don’t know why they are interested in the opportunity. If you fall into this category, this section is for you!
First of all, the easiest way to show interest is to simply explicitly say that you want to work in data science on your resume! Probably the best way to do this is to put a concise objective statement at the top of your resume. There have been many times when I’ve passed on resumes because there isn’t an ounce of indication that they actually want to work in data science. In most of these cases, I would’ve given the resumes a closer look if there was just a short objective statement that talked about wanting to get experience in data science.
Just simply saying that you want to work in data science is a good thing; if you want to do better (which of course you do!) demonstrate how your background is consistent with your data science objective. For example “I’m pursuing a master’s in computer science to acquire the skills to be a top class data science coder.” Or “My background in physics has provided me with the quantitative skills to understand predictive modelling at a deep level.” Not perfect examples, but I hope you get the idea — explaining how your background and academic goals tie to your goal of being a data scientist can pre-emptively answer a lot of questions that I may have while reading your resume — and that is a very good thing!
Of course, words are cheap, and I know that many applicants have multiple resume versions depending on what they are applying to. I generally take the sincerity of an objective statement at face value, but I’ll take it much more seriously if it is backed up by some other things on your resume. Any of the differentiators listed above in tip 1 will work wonders to show that you are very interested in pursuing data science.
Note: Data scientists come from a wide variety of backgrounds. Do not feel that you are at a big disadvantage if you are coming from a ‘non-traditional’ path. We typically will consider any applicant studying a ‘quantitative’ field. But, if your field isn’t directly related to data science you have a little extra work to make sure I understand that you really are passionate about the field.
Tip 3: Realize that no one gets all of the answers right in the interview, don’t get flustered!
I’ve had multiple opportunities to talk to accepted candidates about their interview experiences after they started their internship. Almost all of them say they felt like they didn’t do well in the interview. At first that came as a surprise to me, because (obviously), I thought the were among the best interviewees. I realized they felt this way because they were holding themselves up against a perfect standard — i.e., getting all of the questions right — while I was measuring them up against the other candidates that I had interviewed.
I think it is important to realize that no one does a perfect interview, so if you get a question slightly wrong, or fumble with your words at some point, you aren’t automatically out. I can almost guarantee that your competition has also made some mistakes. I don’t tell you this to make you feel complacent or that the competition isn’t stiff. After all, even if no one does perfectly, you still have to beat most of them to get an offer! My purpose in sharing this insight is to give you confidence during the interview so you don’t ‘crash and burn.’ If you make a mistake or say something awkward, just move on, chances are your competition has also made some mistakes. Try to keep mistakes to a minimum of course, but don’t give up or get flustered because some things didn’t go perfectly! You are probably still in the race!
Tip 4: Do not guess, never guess! Have a strategy for when you don’t know the answer
I think the single worst thing you can do in an interview (barring something totally crazy) is to guess the answer to a question. You might be able to pull it off, and I honestly can’t tell you how many have in my interviews (because I don’t know when some one got one by me 😁!). But I can say if I realize someone is guessing, they lose major points. Not only because they don’t know the answer, but because they are showing really bad judgement. Imagine if someone I hired was later in a meeting with an executive. That executive asks them a question they don’t know the answer and they just guess. What a nightmare!
You need to have a few plans to handle questions you don’t know in a graceful way. Luckily, as an intern you have a solid go-to — “I don’t have a lot of confidence in X, but I will be taking a class next semester that should cover it.” With interns, unlike applicants for full-time positions, we expect you to grow and develop between the application period and your start date. Knowing which classes you are going to take and being able to tie a knowledge gap to one of those classes is about a best-case scenario when you don’t know an answer.
So, saying you will take a class that will cover a specific topic is a great one. But, what if the topic isn’t in your future coursework? You also need a plan for this scenario. If you are familiar with a similar topic, you can say something like “I’m not sure exactly about that, but I do know a general principle about the larger topic is ……” Don’t dismiss your lack of knowledge, or pretend like you didn’t understand the question, but add a little color to let the interviewer know that you are familiar with somewhat similar topics.
You also need to have a plan on what to do if you literally know nothing about the topic. This is a really tough one! My recommendation is to go for honesty and show enthusiasm for learning. Something like this might be the best solution in a worst-case scenario — “I honestly am not familiar with this topic. As a motivated student inside and outside of the classroom, I hope to continue to learn about this and other related topics as I continue to grow my knowledge base.”
In summary, you need to have a plan on how you will answer questions you don’t know the answer to. It should be a multi-faceted plan with a protocol for different scenarios (e.g., you don’t know, but it is a topic you will study later, you’ve heard of it and know just a little bit, you’ve never heard of it in your life etc.). And of course, whatever you do, don’t guess!
Tip 5: Pay attention to and pick up on interviewer hints and cues
Picking up on subtle or not-so-subtle clues and cues from the interviewer can give the interviewer valuable insights on your communication aptitudes. Missing them can be detrimental to your candidacy.
In this section, there are two types of cues I want to address. The first are general social cues and the second are hints and cues specific to solving technical problems.
This is more of general internship/communication advice — just pay attention to the social cues of the interviewers. Listen for polite, but somewhat dismissive ‘uh-huhs’ that indicate that you are talking too long about a subject or you are discussing something that they are not wanting to discuss. I’ve sat through many 15 minute dissertation summaries (in a 45 minute interview) giving dozen’s of ‘uh-huhs’ — which roughly could each be translated to ‘please stop talking.’ If the interviewer is constantly asking you to elaborate, take that as a cue that your next answer should be in more detail. Everyone has different communication and interviewing styles, make sure you are paying enough attention to adapt to the styles and preferences of the person on the other side of the table. Your odds of getting the offer will be much higher!
One question I always ask myself after an interview is ‘Would I like to work on a project with this person?’ Being able to pick up and adapt based on social cues will make the interview a lot more enjoyable for me and will make me a lot more likely to answer ‘Yes!’
I don’t have a whole lot to say about hints and cues about technical questions, except, “use them!” — they are there to help you. If the interviewer tells you to go a different direction in your thinking, then do it! If they tell you to assume something isn’t an issue, assume it! I once asked a question about profit in an interview. The candidate rightly mentioned the importance of discounting cashflows. But discounting cashflows would make the problem too complicated. So I politely told him to ignore discounting for this question. He continued to talk about discounting in all of the follow up questions. He even tried to answer the final big question using discounted cashflows (and he did it incorrectly)! It was a really frustrating experience and he lost major points for his inflexibility.
Tip 6: Be prepared to introduce yourself, be prepared in general
This sounds so basic and honestly, before hiring three cohorts of interns I would’ve scoffed at this overly obvious advice. But my experience suggests this is something that people need to be reminded of. Make sure you have a buttoned up, excellent introduction that is concise and sells all of your strong points. I cannot tell you how many times candidates don’t seem to be prepared for that very predictable question. They often meander, going into too much detail some places and and skipping over detail other places. What a missed opportunity! The ‘tell me about yourself’ request is basically giving you the green light to say why I should hire you. Make sure you have a good pitch!
Less importantly, just make sure you are generally prepared. Know a little bit about the company, have a pen and paper handy, have a copy of your resume to reference etc. basic stuff that won’t win you any points, but will keep you from losing points. And shockingly, there are many people who lose a few points because they seem flustered and unprepared.
Tip 7: Put your expected graduation date clearly on your resume
This is a really simple one. A lot of internship programs target specific graduation dates. If you don’t put your graduation date on there, one of two things could happen (1) you get an interview and they ask for your graduation date there, or (2) your resume gets passed because the hiring manager doesn’t have time to make sure you have an acceptable graduation date. Either way, if your graduation date doesn’t meet the requirements, you won’t get the offer. But if you don’t put it there and your graduation date is acceptable, there is a chance you won’t get an interview because of it. So basically, just put your graduation date. If it isn’t a fit, you won’t waste your time preparing for an interview that will be over as soon as they find out your date. If it does work, you are more likely to get an interview. No reason to not include it!
Tip 8: Understand that the interviewer most likely wants you to do well — they won’t cut you any breaks, but they are probably secretly rooting for you
My last tip is to help you calm your nerves before an interview. I can’t speak for everyone, but when I’m doing an interview, I’m hoping the candidate does well. I don’t want to trip you up with arbitrary tricks or ‘gotcha questions.’ I’m not going to make it easier by any means, but I’m secretly cheering for you. The reason I feel this way is because I want to find great candidates quickly (I’m not a recruiter, I have plenty of data sciencing that I would like to get back to 🤣). I want you to be a great candidate and I’m hoping you are. I think that a lot of interviewers feel the same as way. They won’t cut you any slack, but they are also not looking for tricky ways to eliminate you.
Conclusion
I hope that some of these tips will prove useful to you. Remember that these are just my opinions. Other interviewers and companies may have different opinions and priorities. You definitely want to use your discretion. I do honestly think that following these 8 tips will greatly improve your probability of success in landing that data science internship this year! Good luck!
I’ve hired 3 cohorts of data science interns — here’s my advice on getting an offer was originally published in Towards Data Science on Medium, where people are continuing the conversation by highlighting and responding to this story.
from Datascience in Towards Data Science on Medium https://ift.tt/IBSuFa0
via IFTTT
