Getting your first job as a data scientist can be hard. However, the job market is favorable to this profession, so with enough effort and patience, you will succeed.
Below, we provide the most important activities you should do to increase your chances to be hired as a Data Scientist.
Get on LinkedIn
Looking for a professional employment without having a LinkedIn profile is hard to imagine. If you don't have an account, get it right now and complete all sections. Avoid factual, stylistic and orthographic mistakes.
Don't forget to add to your LinkedIn profile all skills you are good at. Also, ask your ancient colleagues to write recommendations on your LinkedIn profile.
Get on Indeed
Search for data scientist job descriptions on Indeed. Note skills that are most frequently required. Do you miss some of those skills? Learn them using online courses, tutorials, and books.
Apply for data scientist positions, whether you are 100% qualified for it or not.
Build a Portfolio
Your chances of getting the job will be much higher if you have a portfolio of projects where you prove you can work with real data. Your projects (including the code and the blog articles or readme files describing them) can be hosted for free on GitHub.
What kind of projects should you include in your portfolio? Below there are several ideas.
Types of Portfolio Projects
Different projects should demonstrate your different skills and abilities.
A Data Cleaning Project aims to demonstrate that you are cable of taking several noisy datasets, clean them, improve their quality and/or size and combine them to solve a specific problem.
A Data Storytelling Project should demonstrate that you can get insights from data, communicated them clearly and keep your reader engaged.
A Data Visualization Project should prove that you are capable of visualizing data with an appropriate choice of plots and charts.
A Machine Learning Project aims to demonstrate that you are able to build statistical models using arbitrary data, save and load models model, as well as make predictions using the model.
An End to End Project — a project that proves you are capable of building a stand-alone system that can be used in production.
An Explanatory Post in a blog. This post should demonstrate your ability to clearly communicate complex machine learning or statistics concepts to various auditory.
To get a project you could also participate in Kaggle. Put your code and descriptions on Github.
Present Your Portfolio Well
It's not enough to just solve a problem and put your code on GitHub. Each of your projects has to engage the reader. The best way to do so is to write a blog post that "sells" every project, makes the reader feel like you were hired on a contract to do the project. So, first of all, explain what the aim was, then the approach you took and why, then the data you used and the outcome.
You should also make sure your blog post or the GitHub readme file contains instructions on how to install and run your project in case someone wants to reproduce your work.
Don't forget to include all relevant files and data sets. If datasets are too big, you could either provide a link to the dataset or a script that will create the dataset (for example by scraping public sources). For a Python project, you should also provide the requirements.txt file that contains all necessary dependencies to run your code.
Also, add comments to your code, or name your parameters, methods, and classes in a way that it's easy to understand what they do.
Skill-Proving Project: To demonstrate that you have a specific skill, you could also build a project that proves that you are good at it.
Take a Free Online Class
Online classes are a good way to gain knowledge and learn new skills. They also prove to your potential employer that you are an enthusiastic and hard-working person. We recommend to take minimum one class and finish it with a certificate. However, showing three to five courses on your resume would be even better.
Take an Internship in a Lab
Data Scientist's work looks very similar to the work of a research scientist. The employer expects from the candidate to be able to formulate hypotheses, gather data and run an experiment to validate a hypothesis. By taking an internship in a lab, for example in a university or a research institute, you will demonstrate that you have had an exposure to a scientific work.
Look for Different Titles
When you look for job openings online, don't just look for "Data Scientist" jobs. Different companies can call this job using different names. Here are several examples:
- Data Scientist
- Data Mining Specialist
- Machine Learning Engineer
- Machine Learning Developer
- AI Specialist
- AI Consultant
- AI Expert
- Data Mining Engineer
There could be many others. Where possible, search for "Machine Learning" or "Data Analysis" skill and see what kind of job titles come up as a result.
Design Your Online Identity
Your employer will probably search you online, so you have to carefully design what they will find. You can think of creating:
a profile on StackOverflow with multiple given answers and a high score;
a Kaggle profile that mentions the competitions you participated in or won;
A GitHub account with your projects listed and described;
A blog on Medium or your own domain with several well-acclaimed posts;
A Twitter account where you share ideas and links to the relevant material;
Membership in relevant groups on LinkedIn.
Structure Your Resume
Inspire from the resume templates you can download online, but don't follow them religiously. When an interviewer asks "Tell me about yourself" do you give them a chronological account of your life story? No. You start with most important aspects. For resumes that means you lead with your strongest aspect. Maybe it's your education or your job experience or your freelance projects.
Write Ministories, Not Bullet Points
Use STAR Format in Resume
STAR is an abbreviation for Situation, Task, Activity, and Result. STAR assists you in writing short and readable stories about your work, containing all aspects your future employer is looking for.
Keep it Short
We recommend writing one page for 10 years of experience.
Craft Your Personal Branding Message
How can you make sure what you say is what employers want to hear? The trick is the target role. Once you have identified your target role, Data Scientist or Machine Learning Engineer, and have a clear understanding of the ideal candidate's profile, you can then pick your most relevant qualifications and craft them into the personal branding message. For example, "creative data scientist with strong statistical skills and hands-on data science project experience".
Fit Resume to Job Description
Don't send the same resume to all open positions. See what each position's particularity is and update your resume accordingly. You can make a small data analysis project and find clusters of position descriptions and write one version of a resume for each such cluster.
Write the Motivation Letter
Motivation letters are less used today than a couple of decades ago. However, if you don't have a lot of experience that talks for itself, we recommend you to write a motivation letter. It has to explain why you apply for this specific position and why you feel qualified for it.
Only send a motivation letter if it was tailored to a position description. A generic motivation letter would play against you.
Use the same language and document style everywhere. For example, if you use round bullet points, they have to be round and of the same size everywhere. Don't use more than one font face, try to always use the same font size and color.
When communicating with the recruiter boil your communication down to 3–5 sentences that explain:
Why you're interested in the job and company;
Why your skills and background make you a good fit.
Demonstrate your excitement towards the position, show you are passionate. Outwork the competition.
Say "I don't know that" when you don't know something. Don't try to say something just for sake of saying. Honesty is appreciated by most hiring managers. If it's not the case, then anyway such a company is not for you.
Make a Systematic Action Plan
It's often said that job hunting itself is a full-time job. This is true. That being said, without a systematic plan in place, finding a job of your dream might take a lot of time.
Be systematic. First of all, identify the activities you need in job search, and then map them out on a daily basis. Activities might include online application, networking, data science projects, algorithms and data structures practice, interview preparation, and so on.
You also need to define a clear daily or weekly target for these activities. This way, you will be able to make the most out of your limited time for job search, and when everything is done according to the plan, you will be constantly moving closer to your goal.
Use Your Network
Many good positions aren't advertised online, so use your own professional network to find good opportunities. Write to your friends and ancient colleagues and tell them that you look for a Data Scientist job. Ask them to let you know when such a position is open in their companies.
Visit job events and job fairs organized by your city or local university. Many companies have a booth there and you could present yourself to company representatives and leave your resume.
Prepare for Interview
In our blog, we have an excellent article on how to prepare for a machine learning inteview.
Choose Small Company and Low Salary
We advise you to focus on 50–500 sized companies. Such companies usually have senior data scientists and the latter aren't too busy helping and teaching you. Smaller companies also have fewer resources to implement complex hiring pipelines; they also more open to candidates with a limited practical experience.
For your first job, don't ask for a high salary. As a normal salary of a junior/mid-level software engineer. Once you have a couple of years of practical full-time experience as a Data Scientist, you could ask for a raise or find a higher paying job.
Get Feedback on Your Interviews
Proactively seek feedback on your interviews and improve weak areas discovered from this feedback.
Most interviews won't tell you what you're missing so in those cases you need to consider the reasons for why you're not landing offers. It could be that your interviewers think you lack experience or knowledge of problems specific to their immediate needs or you weren't capable of demonstrating expertise through a given task.
For example, if you're getting the "not enough demonstrated practical experience" as a feedback, consider contributing to open source, working on a few more projects and publish your results online.
To truly take advantage of the current high demand for data scientists, consider relocating to a geographical area where employers are less lucky in hiring data scientists. Highly competitive areas like Seattle, San Francisco, San Diego, New York, Boston or Austin are all excellent places to live and they have a lot of great companies competing for talent.
Remember You Are Wanted
The people interviewing you need a new employee as badly as you need this job. They also have a limited time to find a right candidate. So prove them you are the right candidate and they will be happy to stop searching.
You will apply for maybe hundreds of job before getting one. But if you follow our advice, the chances are on your side!
Found a mistyping or an inconsistency in the text? Let us know and we will improve it.