Creating Your Data Science Resume
3 Steps to Developing a Resume that will Stand Out
I’ve been helping aspiring data scientists for 3 years and one of the most common frustrations I hear from those who have attempted to apply to jobs is: “my resume never seems to get a response — what am I doing wrong?”.
Contrary to what many job descriptions might have you believe, a MS/PhD in Stats/CS/Math/Physics and 2–5 years of industry experience are not the end-all-be-all of a great Data Science resume.
So in this article I’ll break down 3 simple steps you can take to develop your resume in a way that will maximize response rates.
Step 1. Make Your Resume Easy for Your Reader to Digest
This may seem like a frivolous detail, but there’s a lot that you can miss out on if you develop a resume that is not easy to navigate or easily readable to the person who receives it. In fact, they may NOT read it at all upon first glance.
Let’s back up a moment and think about who is likely to read your resume.
9 out of 10 times, the first person to read the resume that you submitted through an online application is a recruiter or some other member of the HR team. These are often non-technical people who have been tasked to screen, select, and respond to resumes that they think will be the most relevant in the lot.
Now imagine you’re in their shoes. You’re not a data scientist, and you’ve been tasked to comb through what might wind up being hundreds (if not in the thousands) of resumes over several weeks. Are you going to spend lots of time reading through each and every resume top to bottom — or are you likely to skim and make an informed decision based upon whatever information pops out to you as being relevant to the job description? The latter, right?
This is sometimes referred to as the “6-second test” which stipulates that recruiters may in fact spend no more than 6 seconds on your resume before making an inferred decision to read on and potentially reach out to you.
Because of this, having a resume that is easily skimmed is key. Your resume needs to be attractive to the recruiter by not dominating their time; while giving them the freedom to easily navigate various sections, parse through and absorb necessary information, and not get them bogged down in potentially irrelevant content.
A quick way of thinking about your resume is this — can my resume be read fully in under 1 minute? If the answer is no, you may have too much content.
And remember: you’re applying to an entry-level position. There’s no need for your resume to be bursting at the seems with endless content that documents everything you’ve ever done (and please, do NOT have more than a 1-page resume). Your goal is to showcase relevant experiences, projects, and skills for the given role — nothing more.
Here’s a check list for readability:
- Does your resume include proper “white space” (is the space in-between the margins not overly dominated by text?)
- Is each section easily found and properly labeled?
- Is your resume limited to the most relevant experiences, projects, and skills for the given role?
- Suggested, but optional: Did I use a little bit of color to create visual separation between sections and content?
Step 2. Make it Relevant to THEIR Job Opening
We’ve touched on making your resume relevant, but what does that mean?
An important aspect of relevancy is understanding what the job requires from you; and the job description can be a helpful tool in understanding what type of data science job it is.
When I say what type of data science job, I mean less about what company or industry the job resides in — and more around what data science needs the position requires. For example, below are job descriptions for Data Science roles from two fairly similar life sciences companies in San Francisco — Ancestry & Calico Health.
While the companies are similar, it’s pretty clear that one role cares more about statistical modeling and forecasting via time series data and analysis; while the other cares more about machine learning algorithms to perform text analysis.
So let’s say you’ve done both types of data science work and both exist somewhere on your standard resume. If you’re applying to the Ancestry role, you’ll want to pay extra attention to bringing your NLP work towards the top of your resume (consider the top of your resume to be the most valuable real estate of your resume) so that it has the highest chance to be seen immediately by the recruiter. Perhaps you’ve previously given equal description to each project as well — instead, pay extra attention to elaborating upon the experience that speaks to this role and their needs.
Also, look back at the job description and make sure that your resume reflects those relevant terms and phrases they list. This will make your recruiter’s job easier in being able to quickly identify that the experience you’re showcasing actually closely reflects their needs for this role.
Similarly, you’ll also want to make sure that your “Skills” section expands upon skills that you possess that reflect the needs of the role. For example, perhaps your resume consists of skills such as NLP and Statistical Modeling. But if you apply to the Ancestry role perhaps you’ll want to make sure specific NLP-related skills such as NLTK, LSTM, and Topic Modeling are present — while for the Calico one you’ll want to expand upon Statistical Modeling to include Time Series, Gaussian Distribution, Forecasting, etc. (A quick note about Skills section — while its good to be thorough and comprehensive with your skills as they relate to a given role, never list a skill just for the sake of listing it — you should at least be able to demonstrate a working knowledge of that skill in an interview should you be asked to do so.)
Also, be willing to pass on a role if your skillset and experience don’t match up. For the Calico role, it’s likely you’ll need to be very strong in your quantitative analysis and ability to understand high-level statistical concepts — while the Ancestry role will likely require some pretty advanced machine learning modeling and coding abilities. Are those your strengths? Don’t feel pressured to apply to a role if it sounds like it’s outside of your working knowledge.
3. Make Additional Content Easily Discoverable
As an entry-level Data Science candidate, your previous job may only have a tenuous connection to the world of Data Science. And that’s OK.
However, if you’ve learned Data Science skills on your own, chances are you’ve done some data projects where you aimed to gain insights for a given business problem, and in pursuit of this you embarked on a lengthy journey where you acquired data, cleaned it, did feature engineering, applied a model, got results, and iterated upon that process. (Oh and if you haven’t, I suggest joining us at Metis as we’ll help you build out a portfolio of 5 data science projects using real-world data).
Because of space constraints, your resume can only hold an executive summary of that process in your project (2–4 sentences max). The resume only conveys the gist of what you set out to do, the tools you used, and some preliminary take-aways — but that still doesn’t fully do your project justice.
In addition to the summary, your project should contain a hyperlink to a blog post about your project that details — at length — your motivation, your process, the tools you used, the results you got along the way, your roadblocks, the alternative approaches you took (and the alternative approaches you chose not to take), and it should also include some solid visualizations to help the reader fully understand what you did. (And yes, you should also make your Github code available and it should be in good enough shape for someone to replicate your efforts, with appropriate comments and Read-Me.)
Why all this? Because saying you did a project is nice, but offering your project to a recruiter and hiring manager to review at great length and in great detail demonstrates a working knowledge of the processes that they are looking for candidates to be familiar with. Seeing you write and explain your thought process also gives them a clearer picture of how you think and how you communicate (critical skills for any data scientist).
If you’d like to review a resume that makes project-based content easily discoverable check out one of my alumni Owen’s resume here.
Believe me, if you follow these steps and build your resume so that it makes quality and relevant content easily accessible for a recruiter or hiring manager to find, digest, and understand, you’ll be putting yourself into prime position for a response— even over candidates with those sought-after advanced degrees and multiple years of industry experience.
If you’d like help building your Data Science resume, feel free to get in touch with me at firstname.lastname@example.org or leave me a comment.