Before You Become a Data Scientist

If you’re thinking about a career in data science and don’t know much about it, where to start, or if you even want to make the jump into this world, you’re in the right place.

I began my career as a data scientist after five years of studying a BA in English and a BS in Computer Science. (What??? Yes, I know — this story is for a different time.) As an undergrad, I had actually sworn I would not pursue a career in tech, but somehow found myself scrolling through data science job openings after I graduated from college. But long story short, I was drawn to jobs in which I thought I could further explore the bridge between the humanities and the sciences, where I could apply my story-telling brain to the unraveling of data.

Since then, I’ve learned so much more about this expansive world and am sharing with you what worked for me in helping me land a job as a data scientist as well as what I wish I had known when I began.

So, before you become a data scientist, here are a few things I would urge you to do.

1. Evaluate Your Goals — Why Do You Want to Become a Data Scientist?

I have always believed that people who can balance multiple disciplines and integrate them smoothly into one goal are the ones who can make a true impact in this world. As I considered various job titles that could offer me self-fulfillment, healthy challenges, and an opportunity to explore what my mind could do, I ended up choosing the multi-faceted field of data science. It was an opportunity to combine my affinity for the social sciences, research experience, and technical skills.

My question for you is why do you want to become a data scientist? Is it for the money, the prestige, the novelty? There is, obviously, no right answer for this. But you really need to evaluate what it is that attracts you to this field. The reason being is that “data scientist” is just one of the many titles of professionals who work with data in different ways. When you define your goals for the job, you can make a more informed decision about which job title you’re actually looking for.

If you come from a highly technical background and wish to continue programming all day then you might be more interested in becoming a data engineer, someone who builds the technologies that help bring valuable data to a data science team to evaluate. Whereas if you’re interested in machine learning and statistical methodologies then data science just might be for you. In either case, you’ll be working in the growing field of data science, doing what you do best and learning what it is you want to learn.

If you do think you want to become a data scientist, there are aspects of the job that often aren’t described on the job description but just come with what it takes to build a model. There is a tremendous amount of manual work that a data scientist needs to do in order to understand and clean the data, at least initially. To build a valid test data set, a good data scientist will think through and architect the best way to make sense of the data before applying labels to it. I would argue that I spend most of my time (maybe 80%) researching, labeling data, transforming and cleaning the input, writing documentation, and organizing information. You will be getting nice and cozy with excel, documentation platforms like Confluence, and Pandas (or other data manipulation package).

The machine learning portion of the job is minuscule, and, arguably, plug-and-chug. Until you get to a higher level position with either years of experience or a higher degree in data science under your belt, I’m not sure if you’ll be doing the glorified tasks of a data scientist at your first job. In fact, if you’re really interested in the machine learning then what you’re looking for is a job as a machine learning engineer / scientist. A data scientist will build models using already existing packages and methodologies, not invent new ones.

2. Don’t Be Fooled By Titles — Read The Job Descriptions, and Read Them All!!!

Among the several job titles that are floating around in the market (i.e. data scientist, data engineer, data analyst, etc) some can be misleading. So, please, Please, PLEASE do your homework in reading the entire job description. You will find that the job description of the same job title will vary from company to company. What you should pay attention to are the programming languages (Python, R, SQL, etc) and the technologies (TensorFlow, Tableau, Flask, etc).

A data scientist who is expected to know Python, SQL, and TensorFlow will likely be working with querying data frequently, manipulating it, and building models. A data scientist who is expected to know Tableau and have business or marketing experience might actually be more of an analyst who communicates between the data science team and the business side of the company. By reading a job description in its entirety, you will be able to envision what your day-to-day might actually look like. And, you can ask yourself if the job description matches your WHY.

Sometimes, it’s true that we don’t know our WHY’s. In that case, you should REALLY read the job descriptions. And, may I suggest that you read them ALL? When you don’t know where to begin, or how to define what you’re looking for then read as many job descriptions as you can. You’ll find that some bullet points will stand out to you, and you’ll find yourself feeling disappointed when you don’t see it in another job listing. Or, vice versa, you might see a specific feature of a job such as “acts as liaison between data science team and internal clients”, and will feel a certain aversion to it. In that case, you may learn that you don’t want to work in a client-facing job. This exercise (think of it as research) will help you develop your why or at least tell your why not.

The field of data science is not new but it is new. Many companies are jumping into this now! And — the jobs they are offering will be very different. So, unless the mere presence of the title, “data scientist”, on your official acceptance letter and CV is what gives you total fulfillment, do not simply chase the title. This field is expansive, and there are many opportunities to dive into data science from the angle that is most comfortable to you, whether as an engineer, business person, and/or statistician.

3. Ask the Right Questions — Your Job Interviews Are Just As Much For YOU As They Are For Employers

Candidates often get psyched out by the interview process, focusing so much on whether or not a company wants them that they often forget that these interviews are an opportunity to deepen their understanding of what they are potentially getting themselves into. When they ask, “Do you have any questions for us?”, ask the right questions. I mean, ask the right questions for you.

When you understand you Why and your Why Not, and focus your attention to the job description as opposed to the title, these questions will come more naturally to you. But, of course, I do believe it would help you if I gave you a few examples to show you how to think about these questions, and, consequently, what to make of the answers.

If you come from a background in tech and are more inclined to use certain technologies over the ones listed in the job description, you may want to ask why they choose to use that technology. “I see that you use Luigi, but I am much more familiar with the benefits of AirFlow; Could you tell me why your teams uses Luigi?” Just as an interviewer would like a well thought out answer from you, you might want a well thought out answer from them, especially in this case if you’ve run across many limitations using Luigi, or other software, in the past, yet the company you’re interviewing at is stuck using outdated software. [Disclaimer: I have no bias toward either. Actually, I use Luigi and have never touched AirFlow. This is for the sake of the example. ;) ]

Another great question to ask is about the size of the team, age of it, and existing implementations of what that team has created for the company. Depending on the answer, you’ll get an understanding of how innovative they are, whether or not the team is integrating “data science” to say they’re doing data science, or how many project you might expect to work on in a year. If you want to be in a team that is forward-thinking, ahead of the curve, and produces frequent outcomes then their answer will be telling for you.

4. You Can Do It If It’s What You Want — Do You Have Time, Money, Both, or None?

Let’s say that you’ve read this far and you have absolutely no idea about any of the technical terms I used above, do not fret. Regardless of what background you come from, I believe that data science is a relatively accessible field that is definitely possible to break into. But, the further away your background is from data science, the more motivated you need to be to put in the work to becoming a data scientist.

Though I had my computer science background, I had zero statistics background, which (I didn’t know back then but…) is very important to the job of a data scientist!!! And yet, I landed a job as one. Actually, I landed a job as a Jr. Data Scientist. After my boss interviewed me, he was excited by my curiosity and perspectives, so he wanted to invest in me. Having had little experience in data science, I knew that in order to break into the field, I needed to go back to school to specialize, take some time to build a portfolio, and/or find a job that was willing to teach me. Because I wasn’t sure if data science is something I wanted to invest that much time and money into, I chose the third option to get paid to learn.

If you’re just coming out of a related field, but don’t have direct or extended experience in data science, you can find a position, like mine, where someone wants to teach you, and they’ll pay you for it. What? it sounds like a hoax to you? Well, don’t be too skeptical. There really are many teams out there that, for whatever reason, need a Jr on board who they are willing to teach.

If you’re just coming out of a related field, but have no interest in becoming anyone’s Jr then you need to arm yourself with a CV that can get you a mid-level job. There are many free resources for learning about data science and implementing projects of your own (i.e., Towards Data Science, Kaggle, Harvard, Coursera, etc). If time and money aren’t an issue for you then there are also so many more schools that offer boot camps and masters’ programs that are related to data science. Whichever way you fill your CV, I am sure that your aptitude is what will shine through to companies, and you will land a job. And, guess what, the technologies and methodologies you’ll choose to focus on in your self-assessed curriculum or in the formal class will all be guided by your Why and what you learned from that job description HW assignment.

The jobs you’re interested in will have a set of qualifications. Use them as guidance for how to shape project portfolio and/or academic CV.

The same applies to people who are coming from a completely unrelated background. If you are still in undergrad and have the time and money to make the switch, then go for it. But, if you’re still unsure but want to test the waters then take one of the following types of classes: statistics, introduction to computer science, computer science and business. But if you’re already graduated, or haven’t enrolled in a university then you must work on teaching yourself or getting yourself the degree needed.

Like I said, if money is a problem then that is no problem. A great place to start is by looking at those job descriptions and familiarizing yourself with the technology listed. Google it. Google it all. Google is free. Google is a friend. Seriously. Google, google, google. If any of the technologies interest you then search for tutorials on how to implement them. The resources you need are all out there. Computer science is an extremely accessible field as long as you have a working laptop.

If time is your concern, then still do whichever of the above that relates to you, but actively apply to jobs in the meantime. You seriously never know who will want to take you under their wings.

5. Choose Your Industry and Company Wisely — How Do You Want To Make an Impact?

So, after all that, you know you want to be a data scientist…but what kind of data science? And, I’m not just talking about the technical aspects of the job description any more. I’m talking about the good stuff: the data.

What kind of data do you want to be working with??? And, trust me, this MATTERS. Why? Because this is EVERYTHING you’ll be interacting with… All day, every day. (Side note: This is a great additional question you can spin at an interview.)

How to figure this out? Think about the following…Do you want to be sifting through all of the data related to fashion? education? recruiting? oil? If your current interests align with your background then it would be easy to find an industry through which you can break into data science. For example, say you’re a teacher who’s gone through all of my recommended steps (1–4) and feel that your experience as a teacher will give you a unique perspective and edge as a data scientist working in a company that produces the best learning technologies for kids, then you are probably right! So, use that to your advantage.

Deciding on an industry will either give you the upper hand as it relates to your background and/or become a new frontier you’ve always wanted to explore. Either way, the industry is very important. It will determine the overall experience you have as a data scientist because though the technical responsibilities may be the same, sorting through a bunch of songs will be an entirely different experience from sorting through stocks.

Once you’ve chosen an industry, be mindful of the companies you work for. If simply breaking into the field of data science is important to you then it doesn’t really matter how established the data science team is at your company. However, in my experience, there is a difference between a company that collects data for data science and one that simply had data to begin with. There are many older companies that have so much data they’ve collected over the years but never collected it for the purpose of exploration.

If you’re up for the challenge and prepared to meet many dead ends and loopholes in the messiness of the data then this can be an exciting opportunity for you. But if you’re looking for a company that has data science at the forefront of its products then make sure you adjust your job filter accordingly.

6. Don’t Settle On Your Work Place — What Do You Deserve?

Like I said, the job application process is not just about what companies want. It’s about what you want. And, you deserve it. The idea that you enter a company purely for monetary gain is very outdated. Your well-being as a complete human is extremely important.

So, please, figure out if the environment is what’s healthy and productive for you. Make sure you treat yourself to coworkers and a company that will make you feel like you belong and give you whatever resources and treatment you need. For example, if you are a woman, make sure you’re in a company that not only emphasizes equal opportunity despite gender, but one that actually has at least half, or maybe even majority (or all!) female data scientists. If you are expecting to welcome a child into your family, make sure that your company has the appropriate parental leave for what works for you; even if you’re a dad-to-be, make sure that paternity leave allows you the flexibility you need to spend time with your child.

Let’s say that you really don’t want to be bothered by other people during the work day, make sure your company offers fully remote or flexible work from home accommodations.

But, what if you have no experience? Can you still be picky? You can most definitely be selective. If you know what you want, then ask for it. The company will either say yes or they will say no. And, rejections are the universe’s blessings. You will be the best judge of how to gage what is your priority: money, environment, location, team demographics, etc. And, you can adjust your asks accordingly. Ultimately, you do know your worth. You know what you can offer the company, and the company just needs to decide if that’s what they want or not.

There is no linear formula to finding your way to becoming a data scientist. But there is so much you can do to personalize your experience to ensure that whatever company you land next will be your ideal next step.

I wrote this article so that you are as equipped as possible when trying to navigate the millions of data science roles out there. It can be daunting, but, don’t worry; you will find your way. Whether your pondering over the field turns you toward it or away from it, you will end up where you are meant to be.

Till Next Time, michelle