This article was written as part of the SEMrush Big Blogging Contest
So, you think you can be a data scientist. But, are you sure you have it what it takes to excel in the data science field? Be careful. It’s a very complicated field, and getting competitive day
Data Scientist: One of the Most In-Demand Tech Jobs for 2019
The data scientist job is definitely one of the most lucrative and hyped job roles out there. More and more businesses are becoming data-driven, the world is increasingly becoming more connected and looks like every business will need a data science practice. So, the demand for data scientists is huge. Even better, everyone acknowledges the shortfall of talent in the industry.
But, becoming a data scientist is extremely complicated and competitive. The career path of a data scientist is not going to be easy. It needs a mix of problem-solving, structured thinking, coding, and various technical skills among others to be truly successful.
The Field of Data Science is Broad and Varied
There is no single definition of data science, as it varies with industry, specific business, and what the purpose of the data scientist’s role is. Different roles require different skill sets, therefore the educational and training path is not uniform.
The role the data scientist is to play is now generally broken down into two large categories:
Type A: Data science for people – data collection and analysis to support decision-making based on the evidence
Type B: Data science for software – for example, the recommendations one might get for books or movies from Amazon or Netflix, based upon past behaviors.
Industry Demand from a Modern Data Scientist
In the current job market, Data scientists are expected to know a lot — machine learning, computer science, statistics, mathematics, data visualization, communication, and deep learning.
Being a Data Scientist is much more than a glamorous job title and a generous salary. It takes serious commitment to become a great Data Science practitioner in this competitive, candidate driven market we’re seeing grow exponentially today.
Studying further degrees are extremely time-consuming and take a huge commitment. To then go into the commercial sector, the expectations here can be much more demanding. In a commercial environment, pressure can come from a variety of sources – time, colleagues, money, answers, the list goes on.
Data plays a huge part in business decision-making and the skills required to manage these data sets fall well outside of the remit of managers and executives. This means a lot of pressure can be felt by Data Scientists who are working for companies with shareholders expecting to see profit and business input directly from your insights.
Related: Data Scientist vs Data Engineer
Most organizations will expect some quick results so picking the projects with low hanging fruit becomes important. This can be daunting for a rookie Data Scientist, so some guidance from the wider Data Science team could be key to initial success. Whereas in a research environment, the pressures – whilst still demanding – are perhaps not as pointed.
Required Skills for Data Science Job Roles
Jeff Hale looked at general data science skills and at specific languages and tools separately. He searched job listings on LinkedIn, Indeed, SimplyHired, Monster, and AngelList on October 10, 2018. Here’s a chart showing how many data scientist jobs each website listed.
As per Jeff’s analysis, machine learning, statistics, and computer science skills are the most frequent general data scientist skills sought by employers.
It is interesting that communication is mentioned in nearly half of job listings. After all, data scientists need to be able to communicate insights and work with others.
Among Tech skills, Python is the most in-demand language. The popularity of this open-source language has been widely observed. R is not far behind Python. It once was the primary language for data science. I was surprised to see how in demand it still is. The roots of this open source language are in statistics, and it’s still very popular with statisticians.
Python or R is a must for virtually every data scientist position.
Apart from Python and R, up next are SQL, Hadoop, and Spark. SQL, in particular, is in high demand. It is the primary way to interact with relational databases. Sometimes, SQL is overlooked in the data science world. But it’s a skill worth demonstrating mastery of if you’re planning to hit the job market.
Then come Java, SAS, and Tableau. Tableau – the analytics platform and visualization tool
How to Become a Data Scientist
Choose the Right College Major
In order to become a data scientist, you don’t necessarily need to pursue Bachelors in Data Science. In fact, that’s not recommended at all. It’s actually a bad idea to go for such a niche discipline at the undergraduate level. You could certainly go for a Bachelor degree in Computer Science, Engineering, Economics, Mathematics, Statistics, Actuarial Science, or Finance at the undergraduate level. In fact, you can even go for Natural Sciences or Social Sciences.
Earn a Graduate (MS or PhD) Degree
Enrolling in a Master’s program is the next step, of course. This will provide the theoretical basis. Of course, they are expensive and they take the time to complete. This option, though, will provide a sequenced learning structure and also connections to potential employers who recruit on campuses.
The 365 Data Science study found that 48% and 27% of 1, 001 data scientists hold a Master’s degree and a Ph.D. degree respectively.
MOOCs and Bootcamps
I have been advocating taking online courses for a long time. In order to get a data scientist job, and even for getting admission for an MS Data Science program, self-preparation is very important.
MOOCs are a combination of theoretical and practical approach. If you pick the right sequences of coursework, you can gain the skill sets you need.
Bootcamps are usually taught by practitioners in the field, and the practical experience can be invaluable. These are usually accelerated, and specific projects are built into the curriculum. Sometimes, employment can come from these.
40% of the cohort reported having taken an online course. Additionally, there were 3.33 certificates per LinkedIn profile. So, undoubtedly in order to become a good data scientist, you have to rely on self-preparation by taking various online courses, video tutorials, and MOOC certifications.
Solid Technical Skills
At the end of the day, you need to develop:
- Excellent understanding of machine learning techniques and algorithms, such as k-NN, Naive Bayes, SVM, Decision Forests, etc.
- Experience with common data science toolkits, such as R, Python, Weka, NumPy, MatLab, etc.
- Experience with data visualization tools, such as D3.js, Matplotlib,
- Proficiency in using query languages such as SQL, Hive, Pig.
- Experience with NoSQL databases, such as MongoDB, Cassandra, HBase
- Excellent applied statistics skills, such as distributions, statistical testing, regression, etc.
Refine Soft Skills
Of course, data science is about Mathematics, Programming, and Technology. But, in today’s data-driven workplace, soft skills like excellent communication skills, intellectual curiosity, creativity, cultural intelligence, emotional intelligence, and strong business acumen are equally important.
Related Post: How Soft Skills can Help You to Get a Job
Demonstrate Intellectual Curiosity
Discovery is the ultimate objective of data science. Data Science calls for innovation and creativity in uncovering new ideas. The best data scientists are motivated by their intellectual curiosity to explore data in very creative ways. Top companies are not only looking for people who are good at answering questions, but who want to ask their own questions. A genuine inquisitiveness is rocket fuel for driving a data scientist’s search for meaningful discoveries in data.
Be Proactive & Passionate
Recruiters look for candidates who can demonstrate passion by showing off something they did outside college and work. Take initiative and get involved in a data science project to tackle a real business problem or an investigation. The ability to “think outside the box” and find new solutions to age-old problems differentiates between a great data scientist and a good data scientist.
Interpersonal Skills: Communication & Analytical Skills and Team Work
A good data scientist is literally a middleman between the tech team and the business (strategy, marketing & sales) team of the company. As a data scientist, you need to be a great communicator, a story-teller, and a team player.
You should also put data analytics into perspectives. At times, you need to present the facts and communicate what this means in a way that everyone understands. So, you need solid people skills to drive the company in the right direction when the data implicate for changes in strategies and actions.
To be a data scientist, you’ll need a solid understanding of the industry you’re working in – trends, customers’ pain points, and competitors. You should be aware of what business problems your company is trying to solve. Data Scientists need to know which problems to solve and how to find solutions that work. An understanding of business and being able to relate to changing customer tastes, product cycles and profitability goals are critical to finding truly innovative solution.
Gather Real World Experience
18% of the data scientists reached the top of the data science ladder after completing an internship. So, if you have a Master’s, it is a great idea to look for an internship in the field, rather than going for a Ph.D. straightaway.
In the real world, it would be rare to get employed as a data scientist, right after college. Most of the folks start as analysts (data analyst, BI analyst, business analyst included), scholars, interns, IT specialists, software engineers, and consultants. Only 2% of the folks got their first job as a data scientist.
Top Platforms and Resources to Learn Data Science Skills
The Kaggle survey provided great insights on how to learn cutting-edge data science skills and/or how to become a data scientist. Additionally, the survey also demonstrated the top trends in data science and machine learning across industries.
As you can see from the above figure, the top platforms and resources to learn data science are:
- Kaggle (40% used this resource)
- Online courses (36%)
- Stack Overflow Q&A (34%)
- YouTube Videos (32%)
- Personal Projects (29%)
- Blogs (29%)
- Textbook (25%)
- College/University (20%)
- Arxiv (15%)
- Official documentation (14%)
As per the survey, any data professionals have three platforms/resources (median) to learn data science skills. Professionals with job titles of Data Scientist, Machine Learning Engineer, Predictive Modeller, Researcher or Scientist/Researcher were found to be using four or more platforms. Whereas, the folks with job titles such as Computer Scientist, Data Miner or Programmer used only two platforms.
If you refer to the figure above, you can see that all platforms got pretty good reviews. Most of the platforms received the rating of either very useful or somewhat useful. Below is the snapshot.
- Personal Projects (74% very useful)
- Online courses (70%)
- Stack Overflow Q&A (63%)
- Kaggle (62%)
- Tutoring/Mentoring (58%)
- Textbook (55%)
- College/University (55%)
- Arxiv (55%)
- Official documentation (52%)
- Non-Kaggle online communities (49%)
Career Advice from the Data Science Experts
The article “How to Kickstart Data Science Career” by Mahesh Babu Channa provides very practical advice to start a career in data science for both freshers and seasoned professionals. Among his top tips, he puts a strong emphasis on getting a Master’s degree (MS or MBA) with Business Analytics specialization.
However, he has also put emphasis on building a strong foundation, acquiring strong skills in analytics, reading case studies on data science applications in business, and participating in data science competitions.
Alireza Yazdani also advised learning coding and machine learning in his article titled Tips for Aspiring Data Scientists. He also made a very valid point of excelling at MS-Excel and learning data visualization. I can’t agree on this more. Data is useless if you can’t understand it and/or make others understand. Data visualization is about how to present your data, to the right people, at the right time, in order to enable them to gain insights most effectively. His suggestions on the professional front make this article a must-read.
JT Kostman penned down 25 tips in his article – Advice for New Data Scientists. While reading the opening, I could relate myself as well regarding getting Inmails and queries regarding how to start a career or where to study. JT stressed on being smart, being passionate, being creative, and you must have the heart of a hacker.
I really loved his point on professional degrees – they can act as the ticket to a job. But, once hired, you need to apply your skills. He also advised giving priority to teamwork, continuous learning, presentation skills, and focussing on the fundamentals.
The key takeaways from the post on Top Data Science Skills in 2017 by Lillian Pearson are learning to code in R or Python, know how to use SQL for queries and reforming data, and know implementing machine learning for future predictions. In fact, Peter Eliason also stressed out on the importance of learning SQL in his article. In the modern era, data management tools like NoSQL and Hadoop are more popular. He has got point. To do well in the career you must know both.
Technical skills are no doubt important. But, to stand out in the competitive job market, you need to sharpen your soft skills as well. Know about 5 Essential Non-Technical Skills for Data Scientists by Gaurav Vohra. Kate Strachnyi also wrote about three curated tips in her post Advice for Aspiring Data Scientists. She also lobbied for getting started with the basics and working on real projects before jumping full-fledged.
Job Search Advice by Experts
Advice from Mark Meloon:
If you have something you want the reader or listener [interviewer] to know, you’d better put that up front in your message. For resumes, that means you lead with your strongest aspect. Maybe that’s your education. Maybe it’s your job experience. Don’t feel that you have to follow the order in that resume template you downloaded.
You want to communicate your passion for the field? Do some personal projects. Contribute to open source. Start a blog and be active on LinkedIn.
Advice from Kyle Mckiou:
Turn every bullet point on your resume into a mini story. You’ve probably already got a full page of text, and it’s probably cluttered with one-sentence junk that says “I did this” or “we did that.” Go ahead and delete half of that.
Now that you’ve freed up some space, start expanding on the remaining accomplishments.
Use the STAR format to give each bullet point context and to turn it into a detailed mini story with a resolution.
It’s better to have a few standout stories and accomplishment on your resume than a whole lot of “stuff.”
Also, be excited and passionate and Outwork the competition.
Advice from Eric Weber:
Want to make an impact as a data scientist? Don’t only look at what is being done, but also what is NOT being done. Write out a list of the Top 5 things you could do to help the company. Then pitch your ideas.
1. It is hard to be self-critical. Examining what is not being done is hard but can push you outside the comfort zone of “let’s just keep things running like they are”.
2. Business moves fast. It can be hard to get out of the “get shit done” mentality when things seem to be on fire. But stepping away from that mentality provides a chance to be truly innovative.
3. You know the data really well. Very few others do. Understanding the potential of data is a data science job, not always something that management can always do.
4. Writing out a list forces you to track your thoughts over time. You commit it to paper and it will stick with you. In contrast, just thinking about something doesn’t always make it stick.
5. You must sell your ideas. Simply writing them is okay but without you pursuing your ideas to management, they won’t get off the ground. Pick your favorite one and start identifying its impact and ROI for the company.
Thinking, writing, and selling. Push yourself to do this regularly and you’ll find all sorts of new ideas to share.
Advice from Beau Walker:
Over the past ten years I’ve applied to 898 jobs on LinkedIn. I know this because LinkedIn keeps track. (Thanks for the reminder LinkedIn!)
This number doesn’t include the jobs I’ve applied to on other platforms or directly on employer sites. It also doesn’t include the numerous recruiter emails, InMails, and phone calls I’ve received.
Want to know how many jobs I’ve actually taken as a result of these activities?
Zero. Zilch. Cero. нуль. It’s true.
I get asked a lot about how to find a job. And I talk a lot with people who are discouraged by the application process.
My advice? Consider alternative approaches to finding a job. In the past 10 years, every job I have taken has come from networking. The best jobs often do.
Advice from Vin Vashinta:
When you’re in the mode of answering questions, it’s tough to start asking them. When you’re in the mode to impress, it’s tough to expect the same in return. Remember that hiring is a 2-way street.
Leave an interview with the team wanting more but also expect to leave the interview with the same desire yourself. Were YOU impressed? What did they do to make YOU feel welcome? You’ve put in work to get to where you are now. Gravitate towards those businesses that lift you up rather than diminish all you’ve achieved.
Great companies put in work to leave every candidate blown away, even the ones they don’t hire. Amazon is an excellent example of a company that has impressed me with their hiring process. I’ve had multiple dealings with their recruiters; always professional, quick to respond, & bringing roles that are good fits for my capabilities.
Advice from JT Kostman:
The problem most likely has to do with how you think about your resume.
Most people who get your resume have absolutely no idea what we really do; they just have a list to check. They’re looking for keywords — not concepts. Most of them are not going to be bothered with having to pan for gold. They’re going to give it less than a minute (literally) and move on to the next one on their pile.
Be honest: Is your resume so simple even some Bozo in HR/Recruiting can see how you would be a near ideal match? And is it about you? Or are you clearly showing (not telling) how you would benefit the hiring manager AND the company — including ensuring she can take fewer Tums every day? Are you connecting the dots for HR and drawing them a map?
Advice from Favio Vazquez:
Be patient. You will apply for maybe hundreds of job before getting one (hopefully not).
Prepare. A lot. Not only studying important concepts, programming and answering business questions, also remember that you will be an important piece of the organization, you will deal with different people and situations, be ready to answer questions about how would you behave in different work situations.
Have a portfolio. If you are looking for a serious paid job in data science do some projects with real data. If you can post them on GitHub. Apart from Kaggle competitions, find something that you love or a problem you want to solve and use your knowledge to do it.
The recruiter is your friend. The people interviewing you too. They want you to get in the company, that’s a powerful advice that I remember every day.
Data science is a very broad field and covers a variety of domains from business to bioinformatics. There is no fixed path to becoming a data scientist.
More importantly, the role of Data Scientist has also become a lot more complex. Especially within the field of Artificial Intelligence, Data Science is extremely important; as with the development of self-creating AI software, lower tier tasks are falling subject to automation as Data Scientist’s progress toward more complex duties. The modern Data Scientist needs to be adaptable as the role evolves at the rapid speeds of technological development. Just make sure you’re prepared!
You cannot master everything in one go. You need to start with the basics. More importantly, you must get your hands dirty by implanting your learning on real-world projects.
There is no single/best platform, resource or course. You have to refer to multiple platforms and resources. It’s just like school life – one standard textbook is not enough; you have to read other reference books.
You will come across a lot of advertisements for online courses and graduate (MS) programs. But, believe me, one course/program will never be enough to learn data science. Here are a
Quick Step-by-Step Guide to Become a Data Scientist
- Learn the Fundamentals: Statistics, Data Exploration, and Basic Data Visualization
- Master a programming language; ideally, Python
- Learn the Pre-requisites of Machine Learning: Probability, Calculus, Linear Algebra, and other Machine Learning Basics
- Learn the advanced modules of Machine Learning and Time Series Modelling
- Know how to deal with Unstructured Data (including NLP)
- Get familiar with Deep Learning
- Practice, Practice, and Practice
The main job of a data scientist is coming up with a new meaningful way to interpret the data. So, it’s up to you how to do your job. There is no clear winner between online course (MOOC) and a full-time program (e.g. MS in Data Science). It really depends on your background, existing skillset and career stage.
Even if you get admitted to an MS in Data Science program at a top university, you will need to take a few online courses as well. Similarly, online courses are good to get started. But, getting a few online certifications will not be enough to become a data scientist.
You need to focus on the skills and techniques. An XYZ Certified Data Scientist or MS in Data Science Graduate from the ABC University will not make you stand out in the job market. It’s all about skills and understanding. You can have skills without degrees, and degrees without skills. No matter what, if you are lacking in the understanding and skills, no one can help you.
Additionally, you need to have solid domain knowledge. Domain knowledge can be gained only through real-world experience. So, the key takeaway – brush up your basics, gather quality work experience and keep learning before you think of becoming a data scientist.
Featured Image Source: MarTech Advisor