DiscoverDataScience.org

Finding a Job as a Data Scientist

By Kat Campise, Data Scientist, Ph.D.

Depending on your academic and experiential background, the path to securing a data science job may be as simple as recruiters seeking you out — or you might need to pound the pavement (so to speak) and face a series of “thanks, but no thanks.” It’s certainly the case that, if you peruse data science job postings, you’ll see a distinct trend for the required qualifications: graduate degrees in a STEM field followed by a litany of programming languages and other database/analytics tools. So, before you start applying, the first step is to conduct a self-assessment regarding your current knowledge and skill set.

Are You a Data Analyst or a Data Scientist?

The cross-correlations between the two job functions simultaneously merge and diverge. Data scientists perform analyses, of course. But, they take analysis several steps farther while also dealing the data directly — and not just the nice rectangular Excel spreadsheets; they are pulling, cleaning, and analyzing massive amounts of raw data in preparation for determining which statistical algorithms to apply. This is where the in-depth statistical and mathematical knowledge comes into play.

Have you taken college courses in linear algebra, calculus, and advanced statistics? Furthermore, have you worked with large datasets where you had to apply both the aforementioned mathematical knowledge as well as utilizing Python, R, and SQL (preferably as an advanced practitioner)? Do you have data science projects that you can showcase on GitHub or have you taken a deep dive into a Kaggle competition (or project)? Do you have experience working with Hadoop or large-scale database management systems? If so, then you’ve established a data science foundation.

However, if you’re not quite at this level yet, there is hope: becoming a data analyst can set you up for moving into a data science role. In fact, it would be accurate to state that data analysts are akin to junior data scientists. You’ll still need Python, R, and SQL knowledge (usually at the intermediate level), but the educational requirements aren’t as rigorous: a bachelor’s degree in a STEM field is generally the minimum qualification. In contrast, at the graduate degree level, especially for a PhD., research design and implementation are the focus. Hence, data scientists are research scientists focused on building models, whereas data analysts are using those models — once they are tested and established as robust — to provide actionable information to the enterprise stakeholders.

Target Your Market

Each industry has its own set of business logic, regulations, terminology, and objectives. For example, the financial services industry is likely to focus on machine learning for predicting movements in a particular market, e.g., Forex, commodities, real estate, etc. — or for fraud detection. Meanwhile, the healthcare industry in the U.S. must comply with HIPAA, and understanding DRG codes, the ICD-10, clinical risk adjustment, and health economics research methods are necessary qualifications for a data scientist in that industry. Herein is another point for self-reflection:

  • Do you have any direct education or experience within a specific industry?
  • If you have experience in another industry, where might there be points of intersection between the two different industries?
  • Can you see yourself working with that sector’s data on a daily basis? Or are you more interested in working with several different industries — such as freelancing or contracting work would provide?
  • If you’re interested in a sector for which you have little or no experience, are there courses or certifications you can complete to demonstrate the knowledge?

Preparing Your Resume

One of the hallmarks of a data scientist is their attention to the smallest of details; this characteristic should be clearly displayed on your resume. You’ll need to blend your data science knowledge and skills with the job requirements as described within the job posting. To illustrate, if you’ve performed research in the field of healthcare or have worked in that field, and you’re applying for a data science position in financial services, emphasizing your knowledge of regulatory compliance is a way to bridge the two. The essential skills to highlight are not a mystery — they’re repeated in almost every data science job post and in just about every data science blog or article:

  • Expertise in SQL/Python/R
  • STEM graduate degree or Bachelor’s degree plus x number of years of experience
  • Data mining
  • Familiarity with Big Data platforms
  • Data visualization
  • Good to excellent interpersonal skills (you’ll be communicating your findings to different stakeholders)
  • Understanding of the target industry (g., pharma, real estate, cybersecurity, retail, supply chain management, etc.)

While there is a shortage of data scientists who meet an employer’s job description perfectly, there’s no shortage of people who want to be data scientists – but they have distinct weaknesses in one or more of the required knowledge areas. If you’re applying via Indeed or another job aggregator, you’ll need to get past the human resource gatekeeper (which could be either an actual human or a machine learning algorithm). On another note, make sure your resume isn’t riddled with misspellings or grammatical errors. If you’re not sure where to begin as a reference point for building your data science resume, review the links below:

Alternatively, a professional resume writer can help you put together a more noteworthy resume while also ensuring there are no misspelled words, etc.

Don’t Overlook the Power of LinkedIn

For many, LinkedIn is the digital equivalent of a resume. Recruiters do review profiles and will reach out to potential candidates directly. Grow your LinkedIn network. Make sure you have a professional photo in place. Continue your learning and add academic achievements or personal data science projects to your profile. Consider this your professional social media marketing and networking. You can also hire someone who specializes in LinkedIn profiles to help you. But, either way — if you do it yourself or pay someone — this self-promotional channel shouldn’t be ignored.

When It’s Time to Interview

This is where additional external factors come into play. Each potential employer will have their own process for determining whether you can do what you claim. Many have proof of work components where they’ll expect you to answer statistical questions. These shouldn’t be basic — stats 101 — questions. They may also ask when and how you used a particular statistical method to solve a specific problem. Several sites will give you insight into what might be asked of you, including:

Keep in mind, there is no guarantee that any of the questions listed will be asked. But, it will kick-start your thinking about how to approach that part of the interview.

Freelance Data Science

The gig economy continues to rumble forward with sites such as Upwork, Freelancer, Guru, and goLance connecting millions of freelancers with enterprises of all sizes. You might be surprised to learn which large tech companies regularly hire freelancers for long and short-term projects. There is a caveat: you are running your own business when you’re a freelancer. This means being detailed when you scope your work, setting up payment terms, possibly signing NDAs, finding clients, negotiating contracts, having a contingency plan for scope creep and clients who neglect to pay you. Consequently, your attention is split between managing the contract, managing your business, and doing data science work. If the thought of this thrills you, and you’re ready to jump headfirst into the gig economy, then there’s still more to consider.

Businesses outsourcing data science is a phenomenon that is occurring more slowly than other job types — at least, for the moment. When you compare the number of data science freelance gigs to the demand for freelance content writers, data science has some catching up to do. But, it’s complex work. It’s technical and deeply analytical. Plus, small startups don’t tend to have the budget for a data scientist. Even if they did, they would be wading through profiles of data analysts calling themselves data scientists as a marketing ploy. Freelance platforms are wrought with fake profiles and plagiarised portfolio work. There’s more. If you’re a U.S. based freelancer who has a higher cost of living, you’re competing with international freelancers who may have more price flexibility, i.e., lower living expenses. Add to this that freelance platforms take a percentage of your earnings.

Is freelancing an option? Yes. But, is it a viable option as a data scientist? For those of you have verifiable experience as a data scientist or a graduate degree in STEM, it could be. As long as you’re comfortable with managing your own business in addition to working your data science magic, then start freelancing as a side job. You should continue to sharpen your skill set and apply for a job with an established company. Despite the blogs that proclaim you can go straight into freelancing full time and make tons of cash, this is not true for a majority of freelancers. If you strategize your business carefully, there is the potential for growing into a data science startup. But, it doesn’t magically occur the minute you put together a LinkedIn or an Upwork profile. So, securing a full-time job as a data scientist — or a junior data scientist — will give you the initial stability for transitioning into full time freelancing.