This article is a part of our in-depth Data Science Career Guide. To read the other articles, please refer to the table of contents or the links that follow this post.
Learning data science skills can revolutionize your career. But unfortunately, great jobs don’t simply fall out of the sky as soon as you’ve mastered Python or R, SQL, and the other necessary technical skills. Finding a job takes time and effort. Finding the right job takes time, effort, and knowledge.
The goal of this career guide is to arm you with that knowledge, so you can spend your time efficiently and end up with the data science career you want.
The first step is figuring out what the career you want actually looks like. Where can your new data science skills take your career? Which path is right for you?
Answering these questions should be the first step in your data science job journey. And though the answers might seem obvious, it’s worth taking the time to probe deeper and really explore all of your potential options. That’s what we’ll be doing in this article.
Specifically, we’re going to take a look at some of the different job titles and descriptions that might be options for you if you’re looking to switch careers. We’ll also take a look at options you may not have thought about: going freelance and using data science in your current position.
Switching Careers: What Job Titles Are Available in Data Science?
The first step in any job search is identifying the types of jobs you should be looking for. In the field of data science, this gets complicated quickly, for a couple of reasons:
- There’s no universal definition of “data scientist” or “data analyst” that every company agrees on, so different positions with the same title may require different skill sets
- There are a plethora of other commonly-used job titles that involve data science work that you might not find if you’re just searching for “data analyst” or “data scientist” roles.
Obviously, we can’t cover every potential job title that might be used by a company, but we can talk about some of the major roles in the data science universe, how they differ, and the progression of your career in the field if you’re starting out in that role.
Note: below, we’re using average salary data from Indeed for each position, based on U.S. data. Obviously, salaries will vary by location, company, and based on your own skill set and experience level, so it’s probably best to treat these numbers as rough guidelines. They were last updated on February 8, 2019.
The Big Three: Data Analyst, Data Scientist, and Data Engineer
Average salary: $68,752
What is a data analyst? This is typically considered an “entry-level” position in the data science field, although not all data analysts are junior and salaries can range widely.
A data analyst’s primary job is to look at company or industry data and use it to answer business questions, then communicate those answers to other teams in the company to be acted upon. For example, a data analyst might be asked to look at sales data from a recent marketing campaign to assess its effectiveness and identify strengths and weaknesses. This would involve accessing the data, probably cleaning it, performing some statistical analysis to answer the relevant business questions, and then visualizing and communicating the results.
Over time, data analysts often work with a variety of different teams within a company; you may work on marketing analytics one month, then help the CEO use data to find reasons the company has grown the next. You will typically be given business questions to answer rather than asked to find interesting trends on your own, as data scientists often are, and you’ll generally be tasked with mining insights from data rather than predicting future results with machine learning.
Skills required: Specifics vary from position to position, but in general, if you’re looking for data analyst roles, you’ll want to be comfortable with:
- Intermediate data science programming in either Python or R, including the use of popular packages
- Intermediate SQL queries
- Data cleaning
- Data visualization
- Probability and statistics
- Communicating complex data analysis clearly and understandably to people with no statistics or programming background
Career prospects: Data analyst is a broad term that encompasses a wide variety of positions, so your career path is fairly open-ended. One common next step is to continue building your data science skills — often with a focus on machine learning — and work toward a role as a data scientist. Alternatively, if you’re more interested in software development, data infrastructure, and helping build a complete data pipeline, you could work toward a position as a data engineer. Some data analysts also use their programming skills to transition into more general developer roles.
If you stick with data analysis, many companies hire senior data analysts. At larger companies with data teams, you can also think about working toward management roles if you’re interested in developing management skills.
Average salary: $128,173
What is a data scientist? Data scientists do many of the same things as data analysts, but they also typically build machine learning models to make accurate predictions about the future based on past data. A data scientist often has more freedom to pursue their own ideas and experiment to find interesting patterns and trends in the data that management may not have thought about.
As a data scientist, you might be asked to assess how a change in marketing strategy could affect your company’s bottom line. This would entail a lot of data analysis work (acquiring, cleaning, and visualizing data), but it would also probably require building and training a machine learning model that can make reliable future predictions based on past data.
Skills required: All of the skills required of a data analyst, plus:
- A solid understanding of both supervised and unsupervised machine learning methods
- A strong understanding of statistics and the ability to evaluate statistical models
- More advanced data-science-related programming skills in Python or R, and potentially familiarity with other tools like Apache Spark
Career prospects: If you’re working as a data scientist, your next job title may well be senior data scientist, a position that’ll earn you about $20,000 more per year on average. You might also choose to specialize further in machine learning as a machine learning engineer, which would also bring a pay raise. Or, you can look more toward management with roles like lead data scientist. If you want to maximize earnings, your ultimate goal might be a C-suite role in data — such as chief data officer — although these roles require management skills and may not involve a lot of actual day-to-day work with data.
Average salary: $132,653
What is a data engineer? A data engineer manages a company’s data infrastructure. Their job requires a lot less statistical analysis and a lot more software development and programming skill. At a company with a data team, the data engineer might be responsible for building data pipelines to get the latest sales, marketing, and revenue data to data analysts and scientists quickly and in a usable format. They’re also likely responsible for building and maintaining the infrastructure needed to store and quickly access past data.
Skills required: The skills required for data engineer positions tend to be more focused on software development. Depending on the company you’re looking at, they may also be quite dependent on familiarity with specific technologies that are already part of the company’s stack. But in general, a data engineer needs:
- Advanced programming skills (probably in Python) for working with large datasets and building data pipelines
- Advanced SQL skills and probably familiarity with a system like Postgres
Career prospects: Data engineers can move into more senior engineering positions through continued experience, or use their skills to transition into a variety of other software development specialties. Outside of specialization, there is also the potential to move into management roles, either as the leader of an engineering or data team (or both, although only very large companies are likely to have a sizable data engineering team).
Learn more about the differences between data engineers, data analysts, and data scientists, or take the quick quiz below to figure out which one of these roles might be best for you:
Other Job Titles in Data Science
While data analyst, data scientist, and data engineer broadly describe the different roles data experts can play at a company, there are a variety of other job titles you’ll see that either relate directly to these roles or otherwise involve the use of data science skills. Below, we’ll take a quick look at job titles you might want to consider when looking for employment.
Machine Learning Engineer
Average salary: $144,085
What is a machine learning engineer? There is a lot of overlap between a machine learning engineer and a data scientist. At some companies, this title just means a data scientist who has specialized in machine learning. At other companies, “machine learning engineer” is more of a software engineering role that involves taking a data scientist’s analysis and turning it into deployable software. Although the specifics vary, virtually all machine learning engineer positions will require at least data science programming skills and a pretty advanced knowledge of machine learning techniques.
You may also see positions like this listed as “Machine Learning Specialist,” particularly if the company is looking for a data scientist who has specialized in machine learning rather than a software engineer who can build deployable products that make use of machine learning.
Average salary: $142,049
What is a quantitative analyst? Quantitative analysts, sometimes called “quants”, use advanced statistical analyses to answer questions and make predictions related to finance and risk. Needless to say, most data science programming skills are immensely useful for quantitative analysis, and a solid knowledge of statistics is fundamental to the field. Understanding of machine learning models and how they can be applied to solve financial problems and predict markets is also increasingly common.
Data Warehouse Architect
Average salary: $136,151
What is a data warehouse architect? Essentially, this is a speciality or sub-field within data engineering for folks who’d like to be in charge of a company’s data storage systems. SQL skills are definitely going to be important for a role like this, although you’ll also need a solid command of other tech skills that’ll vary based on the employer’s tech stack. You won’t be hired as a data warehouse architect solely on your data science skills, but the SQL skills and data management knowledge you’ll have from learning data science make it a position that should be on your radar if you’re interested in the data engineering side of the business.
Business Intelligence Analyst
Average salary: $90,150
What is a business intelligence analyst? A business analyst is essentially a data analyst who is focused on analyzing market and business trends. This position sometimes requires familiarity with software-based data analysis tools (like Microsoft Power BI), but many data science skills are also crucial for business intelligence analyst positions, and many of these positions will also require Python or R programming skills.
Average salary: $87,021
What is a statistician? ‘Statistician’ is what data scientists were called before the term ‘data scientist’ existed. Required skills can vary quite a bit by from job to job, but all of them will require a solid understanding of probability and statistics. Programming skills, especially in a statistics-focused language like R, are likely to be of use as well. Unlike data scientists, a statistician will not typically be expected to know how to build and train machine learning models (although they may need to be familiar with the mathematical principles that underlie machine learning models).
Average salary: $78,172
What is a business analyst? ‘Business analyst’ is a pretty generic job title that’s applied to a wide variety of roles, but in the broadest terms, a business analyst helps companies answer questions and solve problems. This doesn’t necessarily involve the use of data science skills, and some business analyst positions don’t require them. But many business analyst jobs do require the analyst to capture, analyze, and make recommendations based on a company’s data, and having data skills would likely make you a more compelling candidate for almost any business analyst role.
Average salary: $73,574
What is a systems analyst? Systems analysts are often tasked with identifying organizational problems, and then planning and overseeing the changes or new systems required to solve those problems. This typically requires programming skill (although systems analysts are not always directly involved in developing the systems they recommend) and data analysis and statistical skills are also frequently necessary for identifying problematic trends and quantifying what’s working well and what isn’t within a company’s tech systems.
Average salary: $66,470
What is a marketing analyst? Marketing analysts look at sales and marketing data to assess and improve the effectiveness of marketing campaigns. In the digital age, these analysts have access to increasingly large amounts of data, particularly at companies that sell digital products, and while there are a variety of software solutions like Google Analytics that can allow for decent analysis without programming skills, an applicant with data science and statistics chops is likely to have a leg up on many other applicants if they also have sufficient domain knowledge in the area of marketing. Plus, a marketing analyst whose analyses make a significant impact can set their long-term sights on a Chief Marketing Officer position, which pays an average of $157,960 per year.
Average salary: $62,468
What is an operations analyst? Operations analysts are typically tasked with examining and streamlining a business’s internal operations. Specific duties and salaries can vary widely, and not all operations analyst positions will make use of data skills, but in many cases, being able to clean, analyze, and visualize data will be important in determining what company systems are working smoothly and what areas might need improvement.
Other Data Science Positions
If you’re searching on job sites (which might not be the best idea; we’ll get to that later), keep in mind that companies use all sorts of titles and that you can adjust any of the above titles to your experience level by tacking words like “junior,” “associate,” “senior,” “lead,” etc. in front of them.
Moreover, these are just some of the traditional full-time career options. If you’re looking for data science work, there are also some alternatives you may not have considered, and we’ll take a look at those now.
Data Science Internships
If you’re looking for on-the-job learning and an entry-level role that’s often a path to a permanent, full-time job, internships are a great option. They’re not for — or even available to — everyone, but they do have some upsides that make them worth considering if you think might be interested in interning:
- They are typically paid positions (the average rate in the US is $20 an hour).
- You get to work with (and learn from) working data analysts and data scientists.
- An internship can easily turn into a full-time position.
- If you have no data science work experience, an internship gets relevant experience onto your resume quickly.
Alyssa Columbus, a Pacific Life data scientist who we interviewed about getting entry-level roles, got her job via an internship, and it’s a path she recommends you don’t rule out. The key, she said, is working hard to exceed expectations during your time as an intern. If you make yourself a valuable member of the team and show a strong interest in learning and growth, you’re a lot more likely to be hired when your internship time runs out.
Of course, there are a few really significant downsides to data science internships that make them difficult for some people to access:
- The pay is comparatively low for the field, and some internships are unpaid.
- Internships typically run for a short period of time (three months is common) and there’s no guarantee of employment at the end.
- Internships are often only available to students, and college-aged applicants may be preferred by some employers.
- It’s difficult to know upfront how much you’ll actually learn from an internship.
For all of these reasons, internships can be a risk, particularly for students who don’t have the financial freedom to take a low-paying job in the hopes that it might turn into a proper data science job later. But if the downsides aren’t deal-breakers for you, then it’s definitely worth considering an internship.
We’ll talk in later chapters about lots of ways you can demonstrate your skills in a job application if you don’t have actual work experience, but if you can get some work experience quickly via an internship, that’s even better!
Going Freelance as a Data Scientist
Although most people who study data science are looking for full-time employment with an established company or startup, it’s worth remembering that data science skills afford you the opportunity to work as a freelancer.
It’s not uncommon that companies have data science work, but not enough of it to justify hiring a full-time data scientist. It’s also not uncommon for companies with a new interest in data science to hire a data science consultant and work through a few freelance projects before committing to permanent data science hires. And of course, even companies with established data science teams may need extra help from time to time. These are all potential clients for a freelance data scientist or data science consultant.
Advantages of Freelancing
You can make more money. Depending on the client and the project, a data scientist with a full suite of skills (like someone who’s gone through most of our Data Scientist path can charge rates of $100 to $200 per hour — or even more. Often, you’ll be able to make more while working fewer hours per week than you might as a salaried employee.
It can take any format you want. You can certainly strike out on your own as a full-time freelancer, but it’s also possible to take on part-time freelance data science work that supplements your regular income, or even just pick up the occasional freelance gig here and there when you’re looking for a little extra cash. Any freelancer will probably need a portfolio site with projects, some information about you, and a list of services, but beyond that, it’s really up to you how much you put into it — it can be as big or as small a commitment as you like.
You decide what you do. Early in your freelance career, you may not have a lot of choice in what projects you take. But once you’ve established yourself as a reliable and skilled freelancer, you’re likely to find you have the freedom to pick and choose the projects or companies you work with.
You decide when you do it. With remote freelance work, you can build your schedule however you see fit. On-site freelance jobs are common in data science and may have prescribed hours, but since you choose which jobs you take, you’ll generally have the freedom to make life choices a salaried employee couldn’t — like working extra jobs over a few months to save up money so you can take a full month off for travel.
Work on a variety of projects with a variety of people. Variety is the spice of life, and working as a freelancer means you’ll be doing different things with different people on all the time. Many freelancers ultimately develop a stable of regular clients, but you’ll be free to switch it up and take on a totally different project or work in a different industry any time you see fit.
This variety can also be extremely beneficial for your career development. Working on a variety of projects will force you to learn and employ new technical skills. Working for a variety of different clients will also help you build some really valuable “soft” skills like communication and client management. If you work across a variety of industries, you’ll also absorb valuable domain knowledge that could benefit you in another freelance job (or full-time employment if you decide to go back).
Downsides of Freelancing
You now have two jobs: one as a data scientist and one as a business manager. It’s easy to forget that while freelance work pays well when you’re actually working, finding work, especially at first requires a lot of unpaid effort. You’ve got to build and maintain a portfolio and website, you’ve got to find and network with potential clients, you’ve got to negotiate project rates, and you’ve got to keep careful track of what you’ve earned and what you’re owed.
You have to be both capable of selling yourself and willing to sell yourself actively. Simply putting up a portfolio site and saying “I’m available” is probably not going to be enough to keep food on the table unless you’re already very well-known in the industry.
Keep in mind that since you’re not a regular employee, you’re often going to be the last thing on your client’s mind. That means you have to put in some extra effort to chase down things you may need, like account or database access. With some clients, you’ll also have to chase down your paychecks (though these are clients you should not work with again).
You can’t count on a stable paycheck. Freelance work doesn’t always flow at a steady rate, and some markets have “seasonal” shifts that may be difficult for you to predict until you’ve been freelancing for a year or two and can start to see the patterns. A company that has a lot of spare fat in its budget for freelancers in the first two quarters of the year, for example, may have a regular expense every Q3 — which means they’ll cut your hours in half. Since you can’t always predict how much you’re going to be able to make each month, working freelance often means you need to build a bigger savings safety net to keep yourself covered.
No health benefits or tax withholding in the U.S. The situation for self-employed people varies from country to country, but in the United States, most freelancers are paid as 1099 contractors — health insurance is not a part of their compensation and taxes are not automatically withheld from paychecks. This isn’t an insurmountable problem by any stretch of the imagination, but it’s one that requires careful thought and budgeting (and setting aside a big chunk of each paycheck for taxes and health insurance). Depending on your individual situation, if you’re going full-time it may make sense to set yourself up as a registered business like a C-corp or S-corp to protect your personal assets from work-related liabilities and in some cases also for tax reasons. You’ll probably want to speak with a local CPA and a lawyer to get a thorough understanding of the regulations and the legal and financial implications of a freelance consulting business based wherever you are located.
You can’t build anything long-term. While the variety of projects you get freelancing can be an advantage, it can also be a downside if you prefer to work on longer-term projects and help them grow and develop over the years. It almost never makes sense to hire freelancers for that kind of work, so you’re likely to get work mostly on shorter-term and one-off projects.
Steep difficulty curve at the outset. Being a successful freelancer is great, but it can be very difficult to get started if you don’t already have a good list of potential clients. Finding solid clients can be a real struggle, particularly if you don’t live in a good local market and have to rely on online remote work, where price competition is fierce. If you’re not sure whether you live in a prime freelancing market, it’s probably best to test the waters first by starting out part-time.
Tips for Data Science Freelancing
If you are going to take the freelance plunge, here are some quick tips:
Consider the upsides and downsides of freelancing sites. Platforms like Upwork, Freelancer, and Fiverr offer easy access to freelance project work, and they can be great places to learn by working through a lot of projects quickly.
It’s important to remember, though, that the convenience they offer comes at a significant cost. First, there’s the direct cost: these platforms take a substantial cut of your earnings. Upwork, for example, currently takes 20% of your first $500 earned from each individual client, and 10% after that. That’s a big chunk, especially when you keep in mind that you probably need to set aside 30% of what ever you earn after that to cover taxes. That means that, for example, when you take on a new Upwork client, the amount that ends up in your pocket after Upwork’s fees and covering your taxes is likely to be half of what you billed, or even less.
There is also an indirect cost to using these sorts of platforms. Because jobs on these sites are easily accessible, you’re competing with the entire world for every project. Every job you bid at a reasonable rate for your location and skills is likely to get quite a few lowball bids that promise the same results as you for a fifth of the price. There are workers on Upwork, for example, who claim to be data scientists but charge less than $10 per hour. Can they deliver the same kind of quality as you? Probably not, but you’ve got to be a pretty skilled salesperson to convince clients to accept your much-higher bid on a regular basis.
Additionally, there’s an inherent risk to any freelance business that relies on any third-party platform, because the platform could change its rules, suspend or delete your account, or simply cease operations at any time. Users typically have no control or influence over platform-wide policies, and changes to these policies can dramatically impact your business.
Remember: you don’t have to use any of these platforms to be a successful freelancer. Although it requires more up-front effort, developing local and regional clients through real-world networking may pay more dividends in the long term. This approach will allow you to build more reliable client relationships, and bill at a level commensurate with your skills and with the cost of living in your region. It is also less risky, because the existence of your business isn’t contingent on the existence of a third-party platform you don’t control.
Offer a clear list of services and bill based on that. While you can bill by the hour, you can often make better margins charging a per-service rate. This also helps ensure that you aren’t stuck doing busy work or boring tasks unrelated to the services you offer simply because your client purchased 10 hours but you finished the project in eight. Having a clear list of services sets better expectations on both ends: you know exactly what you have to do and what you’ll be paid to do it, and the client knows exactly what they’re getting and what it will cost them.
Start small. Making the jump into full-time freelancing is less of a risk if you’ve already been doing it part-time and have a roster of regular clients. Starting with some part-time or side freelance work will also help you develop the organizational skills and workflows you’ll need to manage the business side of being a full-time freelancer. While it’s possible to go straight from full-time employed to full-time freelancer without any previous freelance experience, you can avoid a lot of stress and struggle by starting small, and doing that will give you a chance to test the market in your area and see how much you really like freelance work before you take the full-time self-employment plunge.
Getting a Raise at Your Current Job Using Data Science
Finally, it’s worth pointing out that data science probably offers benefits that can help you in your current career even if you have no interest in becoming a full-time or freelance data scientist. Precisely what you can do will depend a lot on what your job is, but if you’ve got some data analysis skills, you’re almost always going to be able to add value in some way. And if your data analysis skills can make a dent in the company’s bottom line or improve your own productivity, then they could help you earn more in your current role.
Consider, for example, Dataquest student Curtly Critchlow. At the time we spoke with him, Curtly worked at the Livestock Development Authority in Guyana, and part of his job involved working with spreadsheets (as many jobs do). Because his department produced a lot of data, this monthly Excel task became a week-long nightmare — until Curtly learned some data science programming skills and was able to turn that into a task that took him just a few minutes.
Imagine how much more you could do with an extra week each month!
Not every example will be so dramatic, but there are ways that data analysis can increase efficiency in almost any job.
If you’re looking for data science opportunities in your current position, there are two easy places to start:
First, look for places you can save time or increase efficiency by applying data science techniques. Your team might already be doing effective data analysis in Excel, but could that process get faster and more easily repeatable if you applied your Python or R programming skills?
Second, look for existing data sources at your company that are being ignored or under-used. This sort of thing is common, particularly at companies without a data science team. Maybe it’s ignored because no one has time to dig into it using the inefficient data analysis methods they know. Maybe it’s under-used because not enough people know how to do any kind of data analysis. In either case, applying some real data science skills can provide significant (and often unexpected) value to your company.
Third, look for ways to optimize your own performance. In the age of personal data-tracking gadgets like smartwatches, it’s quite possible to track and analyze your own data in ways that can make you more effective and productive. Here are a few examples of cool things you can do with just some basic programming skills (the article focuses on R, but the same things are also possible with Python). If you start poking around, you’ll find that a lot of the platforms you use for both life and work allow you to export data, download CSVs, or otherwise access your own data for some personalized number-crunching.
Am I Ready for a Data Science Job?
The easiest way to assess your own readiness is simply to start taking a look at real-world jobs and job descriptions. Do you have the skills that are listed there? Do you feel like you’d be able to do (or learn to do) the tasks described?
Your answer to these questions doesn’t have to be a rock-solid yes. Impostor syndrome is a real thing (here are some tips for combating it), and particularly for entry-level applicants searching for their first data science job are particularly susceptible to feeling it. It’s easy to look at an employer’s wish-list of skills and qualifications and intimidate yourself out of even applying.
When we talk to former Dataquest students with full-time jobs in data science, they regularly advise that other students apply for jobs even when they don’t feel ready. Amazon data scientist Caitlin Whitlock, for example, says she the prospect of her interview at Amazon was “terrifying.” But she still advises that aspiring data scientists “apply for any job, period. If you don’t think you’re going to get it, apply anyway.”
Miguel Couto, who got three job offers after applying for jobs on a whim, before he thought he was truly ready, agrees. That doesn’t mean you should go in unprepared—both of these students also said that they prepared really thoroughly for job interviews—but it does mean that you might be ready to get a data science job before you actually feel ready.
Do I need a certificate?
Whether or not you need some kind of certification to get a job in data science is another common question that students just starting to think about the job application process routinely ask. The short answer to this question is: no, you do not.
In fact, none of the recruiters and hiring managers we interviewed for this guide mentioned certificates as important, or talked about using them to assess data science applications. We asked every interviewee what made data science applicants stand out in terms of both resumes and in interviews, and not a single one of them mentioned certificates even once.
Certificates do have some use, in that they can help demonstrate you’re committed to learning, and actively working on improving your skills. But there’s no must-have certificate for data science, and it’s very, very unlikely that any certificate would be the thing that convinces a recruiter to hire you or even to give you an interview. Even the certificates from brand-name colleges aren’t very useful in that regard, because hiring managers know that these programs are often administered separately from the university’s regular operations, and standards for passing are often quite lax.
So when you’re thinking about whether you’re ready to apply, don’t worry about what certificates you have. If you do have certificates, that’s great, but if you don’t, you certainly don’t need to rush out and try to get one. There is no must-have data science certificate. What really matters for getting a job in data science is your skills.
If you need those skills, Dataquest can teach you! But since this guide is focused primarily on finding jobs in data science, let’s assume you're all skilled up and move on to the next step. Once you’ve identified the kind of job you want, where can you actually find it?
This article is part of our in-depth Data Science Career Guide.
- Introduction and Table of Contents
- Before You Apply: Considering Your Options — You are here.
- How and Where to Find Data Science Jobs
- How to Write a Data Science Resume
- How to Create a Data Science Project Portfolio
- How to Fill in Application Forms, When to Apply, and Other Considerations
- Preparing for Job Interviews in Data Science
- Assessing and Negotiating Job Offers