Learning the required skills can sometimes be considered the easy part of the process of becoming a data scientist. The hard part often comes later when it comes time to apply for jobs.
Nowadays, many people can call themselves data scientists despite varying competencies and abilities to provide impact, which means that you need to stand out from the crowd.
With modern job application processes, however, it can be hard to even get a foot in the door when you can’t talk to a recruiter in person. With many of us sending our resumes off into the void, it’s difficult for us to know whether or not we’re making the right impression.
Luckily, there are some easy ways to spruce up your resume to ensure that recruiters are giving it a longer look than the average 7.4 seconds.
Here are five different things in your resume that are keeping you from getting a job in data science, and five ways you can fix them:
1. Skill-level bars to indicate skill proficiency
One of the latest trends in resume-building is a graphical representation of skills using skill bars, percentages, or other graphics to demonstrate your level of skill mastery.
While this option is available on many resume templates intended to be sent to hip, startup-style companies, it’s doing you a disservice and potentially discounting you from the job application process before they’ve even met you.
First, skill-level bars are utterly meaningless to a recruiter. Unless you develop a point scale system, how does a recruiter know what "4/5 stars in Python proficiency" means? Additionally, to help the recruiter better understand your rating system, you would need to develop criteria and objective tests for each skill so you can explain why you would give yourself the score you did.
Second, skill bars are subjective. In short, people are bad at rating themselves on anything, with more intelligent people tending to rate themselves lower on skills they are proficient in and less intelligent people rating themselves higher on skills in which they are not proficient.
Third, your ability to rate your skills is completely unreliable. Without a reliable, widely-known testing scale you can use to determine your skill proficiency, such as the Common European Framework of Reference for Languages which is recognized the world over, your own ability to ascertain your skill level is entirely unreliable. For example, if you tell someone that you are at the B2 level in Spanish, they will likely understand what your proficiency in the language is. However, if you tell someone that you know 80% of Python, they’ll have no idea what your abilities are – because, let’s face it, what does an 80% knowledge of Python look like?
In the end, using skill-level bars on your resume discounts you from many job opportunities before you even get a foot in the door. A recruiter may see that you are an ideal candidate, but they may prefer to interview someone who has an 80% proficiency in Python compared to your 70%.
What to do instead
Instead of skill-level bars, discuss the skills you used in your description of previous employment and projects to demonstrate how long you’ve been using the skill, what you’ve used the skill for, and what impact you’ve made by using the skill.
2. Lack of personal projects that demonstrate real-world problem solving
One of the most common complaints of tech recruiters is that candidates lack real-world problem-solving skills.
It’s easy to solve a problem in university when it’s laid out nicely in front of you with a beginning, middle, and end, and a professor who is ready to answer your questions about how to get the desired result.
However, the real world will throw problems at you that have no concrete beginning, middle, or end, and may only have a vague business question that is now your job to answer using data. The working environment that you’re applying to be a part of isn’t as squeaky clean as what you experience in university or coding bootcamps, which means that you need to develop your real-world problem-solving skills.
What to do instead
The key here is to remember the old saying, "show, don’t tell."
While it’s easy to tell an employer that you’ve solved many real-world problems through your independent research projects, it’s much more impactful if you show them the problem you solved and exactly how you solved it.
A personal portfolio linked to your resume will help you out here. In that portfolio, you want to develop several projects that figure out specific problems within your target industry. For example, if you want to work in climate science, fill your portfolio with projects that analyze air pollution levels and describe how pollution levels correlate to hospitalizations in contrasting major cities.
3. The wrong set of programming skills for the job
First-time data science job seekers will set out with a resume listing 20 programming languages they are proficient in without focusing on the languages that an employer is specifically looking for in their job ad.
Listing 20 programming languages or skills is useless to an employer who only wants to know if you can work with Python, Excel, and SQL. Additionally, this shows a lack of preparation on the job seeker’s part because it presents a resume designed to be a "one-size-fits-all" solution to the job application process.
Remember, more isn’t always better. A recruiter would rather see concrete evidence of experience with the programming languages and skills their company is looking for than random experience in 20 different languages.
Recruiters tend to only look at resumes for 7.4 seconds, which means that if they have to sift through your list of skills for the ones that apply to their company, then you’ve already made yourself an easy candidate to weed out.
What to do instead
Prepare a tailored resume for each job application by carefully scanning the job ad and highlighting the exact languages and skills they are looking for. Then, make it easy for the recruiter by giving concrete examples of your ability with that language or skill, and link to projects that demonstrate your abilities (again, "show, don’t tell").
It’s important to focus on the languages and skills a specific company is looking for because their business processes revolve around those languages. Unless a company is hiring you to set up their data analytics processes, they’re looking for a candidate who can meld seamlessly into their existing teams and processes using their required languages and skills.
In short, the focus should be on making it easy for a recruiter to visualize you successfully becoming a valuable member of the team.
4. A focus on algorithms and code and not outcomes
As so succinctly stated in Ken Jee’s article Data Science Resume Mistakes to Avoid, the value of data scientists is their impact on a company and the outcomes their work produces.
If your resume focuses more on the fancy algorithm you used to solve a problem instead of the outcome of solving that problem, a recruiter may not be convinced of the impact your analysis had.
What to do instead
Instead of focusing the entirety of your project description on the algorithms you used or the code you wrote, add sentences where you describe the impact your analysis had or the outcome of your project.
This description should include quantifiable information, such as percentages, growth indexes, and monetary values. For example, you could say that the pro-bono project you did for a small business in your town helped them increase the efficiency of their processes such that they increased their revenue immediately by 30%. Alternatively, you could say that your analysis of atmospheric pollution levels has indicated that global temperatures could rise 2 degrees Celsius by 2100.
Giving a recruiter concrete details about the impact your project had or the outcome it produced will give them a sense of your ability to impact their company positively.
5. No indication of lifelong learning
As mentioned previously, a data scientist is only as valuable as the impact they provide to a company. This means that data scientists must remain relevant throughout their careers by constantly learning new skills, educating themselves about new trends in the sector they work in, and exposing themselves to the latest technology that could prove more efficient for the company’s processes.
No indication of continued education, whether through courses, bootcamps, or personal projects, or a large gap between the current time and the last course you took, could indicate to a recruiter that you’re not serious about the lifelong learning that is so critical to data scientists.
What to do instead
Make a point of working on personal research projects every so often to show recruiters that you are always looking to improve your skills and learn something new.
Each new project doesn’t have to be huge, but it should integrate some new technology, algorithm, or thought process to tackle a real-world problem. Each project should be described in your resume using a short sentence that captures the advancement of your skillset and the problem you set out to solve.
Adding links to your project repository or portfolio is a great way to give recruiters a hands-on look at your work.
Key takeaways
- Don’t use skill-level bars to indicate skill proficiency because they’re meaningless to recruiters and because they are unreliable indications of your competency in a given skill.
- A personal portfolio linked to your resume can give recruiters a hands-on look at your work and can give them a sense of your ability to solve real-world problems.
- Focus your resume around the skills and programming languages required for each unique job.
- Use quantifiable variables to indicate the impact and outcome of your projects.
- Keep educating yourself throughout your job-seeking process and the rest of your career to remain relevant – each new project you complete for your portfolio should integrate new skills to indicate a commitment to life-long learning to recruiters.
Subscribe to get my stories sent directly to your inbox: Story Subscription
Please become a member to get unlimited access to Medium using my referral link (I will receive a small commission at no extra cost to you): Medium Membership
