Is this a data scientist? Asks an HR person before attempting to write a data science job description.

15 Data Science Job Description Tips: How to Write a Great Job Posting

Is the Great Resignation leaving you desperate to fill some vital data science roles in your organization? Before you send out the same old job posting as last time, learn how to optimize your data science job description to attract the right candidates.  

As data science recruiters, it’s our job to connect employers with outstanding candidates for their open data science, engineering, and analyst positions. We regularly help our clients tune up vague or ineffective job descriptions so they really appeal to the data talent who best fit their needs and culture.

It’s no secret that job descriptions for data science seem to be rife with issues. And job descriptions, in general, are just pretty dry. So often, they’re little more than a list of requirements and duties. But there’s a lot of nuance in how you present and structure what you’re looking for in a data scientist or engineer that can have a big impact on how well your job posting attracts applicants. 

We’ve written before about how to make your company more appealing to data science candidates. Here’s how to translate that effort into a great data science job description.

Before you start to write a data science job description: 

  • Identify your goals for the job posting. Usually these are along the lines of attracting the right data science applicants and discouraging the wrong ones. But you also want to consider how the job posting will reflect on your company. Are you all-business go-getters? Or a super-chill, culture-focused team with free lunches? Or is it somewhere in between? Your tone, choice of words, and amount of info you choose to disclose (such as a salary range—more on that later) says a lot about who you are as a company. 
  • Make sure the role, itself, is appealing to the type of data talent you want. Most highly-qualified data scientists and engineers are going to want opportunities for growth. They want to see that there is a career path ahead of them within your organization (or elsewhere). They want to know how they’ll expand their experience and skills set working with your company.
  • Visualize the perfect candidate. Do a bit of candidate persona development on what your ideal applicant looks like. Ask yourself: What can they do? What have they done? What do they want? Consider things like goals, challenges, lifestyle, and location. How can your job help them attain these things? For instance, we sometimes find it effective to try to appeal to job seekers who’ve left a geographic area but would love the chance to return. 
  • Set a competitive salary. Do this up front and be prepared to be transparent about it—don’t wait to negotiate the lowest rate you can get at the end. This is not how you attract top talent! Competitive pay is a major selling point for a data science job (any job, really). If your company can’t offer the going rate for a data science position, you’ll need to consider how you can sweeten the pot with benefits. The data talent market is simply too competitive to underpay and expect high-quality candidates. And we’re not just talking about things like healthcare or 401Ks. Consider perks that have a real impact on everyday life, things like fully remote work options, flexible schedules, paid training, or extra time off. We’ve seen candidates reconsider a job with less than desired pay if there’s a reasonable trade off in perceived happiness.

Write the data science job description

  • Write an accurate job title. We see this all of the time: employers will post a job for something like a “Senior Data Scientist,” when what they really need is a data engineer. Additionally, don’t call a position “lead” or “senior” if it’s not, but don’t leave it off if it is! Another thing to consider is how common is the terminology you’re using? For example, you may have an opening for a “Growth Analyst” but if the role is largely what other companies would call a “Data Scientist,” you might want to combine the terms. Job titles hold lots of meaning for people and can be a significant selling point. We’ve had applicants turn down a job offer because they didn’t like the title.
  • Include the specifics. No, it’s not always given that a job is 40 hours a week, or on-location only, or has 9 to 5 working hours. There are a lot of new rules for our global employment market—you need to be clear with your expectations for a particular position. Include things like: full time or part time; contract or permanent; required working hours and if they are flexible; remote and acceptable time zones for remote candidates, on-site or hybrid work site; travel requirements and how much; etc.
  • Note any immigration limitations you have. But be conscientious with your wording, as it can have legal/discriminatory implications. For example, we simply state, “sorry but we cannot sponsor visas at this time.” It’s important to consider, however, that much of the technical labor force has immigrated to the U.S. If you are able to sponsor visas, it can be a significant  differentiator when competing for candidates.
  • Specify the salary range. We already mentioned that you should determine your salary up front (see #4). Now it’s time to put it in the job posting. Job seekers want to see a salary range (and listing benefits doesn’t hurt, either). When they don’t, the message is that you’re just trying to hire as cheaply as possible. Not to mention, when you’re reluctant to disclose pay up front, it can waste everyone’s time. It’s frustrating for everybody for good candidates to move all the way through the application and interview process only to discover that the role pays less than what they currently make. It’s sometimes helpful to note that there is some salary flexibility, within reason of course, for uniquely strong candidates.
  • List actual job requirements. Be clear about what you truly need, but don’t overdo it, either. Data science job postings are well-known for their ridiculous requirements. All this does is show candidates that whoever wrote the job description has little real-world experience in data science (and didn’t bother to ask). And this doesn’t instill confidence that your company is a great one for advancing their career. Our advice on job requirements: stick to the deal breakers. If you really want to include some nice-to-have skills, list it separately as “preferred qualifications.” But again, don’t overdo it.Well-thought, clear, and concise job requirements have the added benefit of helping candidates identify when they don\’t meet the mandatory experience. If job seekers see an unrealistic laundry list of requirements, they’re more likely to apply even if they only have a few skills, because it’s not abundantly clear which ones are truly necessary (and frankly, they assume you didn\’t know what you were doing when you wrote the job description). As you can imagine, this only serves to waste everyone’s time.
  • Leave out superfluous requirements. These are things, such as, “good communications skills,” “professional attitude,” and “able to deliver work on time.” What job seeker is going to say that they don\’t have these skills? (And what harm is it, really, if you leave it off? Is some job seeker going to say to themselves,  “Woo hoo! I found the one job that doesn’t require ‘good attention to detail’—I’m applying for that one!”) A much better route, and the one we use ourselves, is to develop interview questions that help identify these types of traits.
  • Be conscious of requirements that are discriminatory. Statements like “looking for a young, go-getter,” or use of gendered pronouns are not appropriate for job postings.
  • Don’t focus too much on years of experience. Data science is a relatively new and constantly-evolving field. Many tools and technologies that have widespread use haven’t been around all that long. We see lots of simple mistakes in this respect in data science job descriptions—employers who require something like five years of experience in a technology that was only released three years ago.

    More importantly, when it comes to years of experience, remember that some people are enthusiastic naturals and others are just folks who need a job. So it\’s not unusual to have a data analyst with two years experience who\’s actually a better candidate than the one with five. Yes, field experience is important but it\’s not the years that really count here, it\’s the competence and the ability to react independently and knowledgeably when unexpected situations occur.
  • Provide context for the position, not just duties. Job seekers like to know where the job falls in the overall organizational structure of a company; things like who they will report to and/or oversee. It’s also helpful to specify the balance of hands-on technical work vs managerial/strategic functions. This kind of detail is useful since job titles (and even adjectives like “senior,” “lead,” “principal,” etc.) can mean very different things in different organizations. Candidates want to know whether this role will be an upward move, a lateral move into a new industry, or a shift in responsibilities.
  • Sell the job. The market for talent is a competitive one—be a salesperson! Put your job in terms that your perfect data science candidate (see #3) would find hard to resist. How much better could life be if they’re in this position:
    • How will their work make a difference? What cool and important problems will they be solving? How will it help them keep their skills on the cutting edge?
    • Where can they go from here? What career opportunities does this job open that others do not?
    • Why is your organization a great one to work for? How does the company’s mission contribute to more than just it’s bottom line?
    • What’s the company culture like? What does it feel like to be part of the team here?
    • What does the location bring to their lifestyle? Does it provide an exciting cosmopolitan environment? A quiet, affordable, and safe community? Lots of natural beauty and opportunity for outdoor adventure?
  • Clearly state how to apply. We know it seems obvious, but it’s easy to forget when you’re focusing on all the other details. And it’s an added bonus for the job seeker if you can make it as easy to apply as possible. 

A great data science job description = great applicants! (or == if you’re a Python developer)

The first step to getting the type of candidates you want for your data science, analytics, or data engineering position, is to write an effective and compelling job description. You may be in a rush to get the job “out there,” but it’s worth every minute of your time to craft an appealing case for working with your company. You’ll thank yourself in the long run when you spend less time wading through a pool of poorly-qualified candidates and are able to more quickly hire the right data professional for your company.

Do you need help hiring great data scientists?

We’re experts at matching employers with the right data scientists, data engineers, and analysts for their specific needs. Contact us today to learn more. 

Scroll to Top