Table of Contents

• [toc headings="h2" title="Table of Contents"] In the not too distant past, big data was the domain of technology organizations. That is no longer the case, and companies across industries today rely on data to drive their decision-making and day-to-day operations. As a result, the role of data scientist is increasingly in demand, with the number of jobs in this field expected to grow by 35% through 2032, with an average of nearly 18,000 jobs being added each year.  Already, this has led to a shortage of job applicants with data science qualifications, and that is an issue that is unlikely to resolve in the near future. In the Adastra 2024 Data Professionals Market Survey, 76% of US data professionals believe that the current talent shortage will continue at least through the end of 2024. They specifically cite a widening talent gap in Senior Data Analyst positions that can oversee an organization's end-to-end data framework, as well as in frontline talent that can help companies integrate and utilize new tools and strategies.  For companies that need to hire data experts, doing so isn't likely to get any easier in the next year. Using a strategic hiring approach can help you find and attract the limited talent in this active market, ensuring you can get the full benefits of data-driven decision making on your team. 

• Understanding the role of a data scientist

• Data scientists have quickly become key players on an organization's business intelligence team. These professionals collect and analyze large volumes of information, with the goal of finding solutions to business problems or providing insights that can support the company's future growth. To do this, they use a combination of programming skills, machine learning algorithms, and in-depth knowledge of mathematics, statistics, and data analysis and modeling strategies.  The everyday work of a data scientist depends on the needs, size, industry, and nature of the organization and its projects. In some cases, they will be part of a broader data science team, working alongside data analysts, machine learning engineers, statisticians, and other data professionals to gather and process data and identify patterns in it. In a smaller team, a data scientist's role is often broader, encompassing the entire range of tasks involved in turning raw data into actionable insights. Whatever the setting or scope of the position, however, the key responsibilities of a data scientist are the same: to use their analytics expertise to deliver information that others in the business can use to make better decisions. 

• When do you need a data scientist?

• The short answer to this question is that having a data scientist on your team is an advantage in any data-driven business. Of course, defining just what qualifies as "data driven" has gotten more complicated as the volume and variety of data increases. Even otherwise low-tech businesses today often use data from customers to shape their marketing content and strategy.  Something else to keep in mind: just because your company uses data doesn't necessarily mean you need to hire dedicated data science professionals. There are a number of programs, products, and resources that people can use to apply data science methods to solve a business problem. Even if you do require data experts, a permanent full-time hire isn't always the best solution. In some cases, it will be more cost-effective to utilize third-party data science services from an agency, particularly if you only have a part-time or occasional need for a data analyst.  So when do you need to have in-house data employees? The answer to that question will be different for each business, but there are some signs that it may be time to hire data team members. The first is if you have ongoing data science projects that are integral to your business. In-house data science teams combine a first-hand understanding of your field, mission, and company culture that a freelancer or third-party agency lacks, which often results in higher quality data analysis as well as a more efficient data analysis process.  The industry that you're in, and the size and scope of the company, will also be factors when deciding whether to hire a data professional. If your work is focused on technology and you have daily work related to data visualization and analysis, then it will likely be worth it for you to create dedicated positions in that area. 

• Common challenges in hiring data scientists

• The high demand for data science professionals is the biggest challenge most employers face when hiring into these roles. There aren't as many qualified data science candidates as there are jobs for them, creating a very competitive hiring market. An experienced candidate will often get multiple offers. Smaller organizations that lack the name recognition or payroll resources of large corporations may struggle to attract candidates as a result, particularly since the top talent typically commands a high salary. Even entry-level data scientists earn an average salary of $92,000 per year, and that pay expectation rises for professionals with a more established background in the field.  Simply identifying what kind of data scientist your organization needs can be its own challenge. Some roles in this field demand specialized skills in areas like natural language processing, AI, deep learning, or business intelligence. In other roles, a professional with a broader skillset will make more of an impact. Since those who make hiring decisions often don't have expertise in these technologies, it can be difficult for them to accurately assess which skillsets are the most important in a candidate.  The broad scope of the data science field raises additional questions for hiring teams. Data scientist is often an interdisciplinary role, requiring expertise in topics like strategic decision-making, business intelligence, and storytelling, in addition to their proficiency in math and computer science. In addition, it's often beneficial if they have work experience in your business domain, particularly for industries like healthcare and finance.  The bottom line is that finding and hiring the right data scientist isn't easy and requires a different approach than the typical talent search. This may mean proactively headhunting passive candidates or reaching out to professionals directly on LinkedIn or a similar social media platform rather than waiting for applicants to respond to your job post organically. In other cases, it can require seeking out new sources of talent, for example by advertising entry-level positions at universities to attract recent graduates, or by utilizing a recruitment firm to gain access to their network.

• How to hire a data scientist step-by-step

• Step 1: Define the requirements of the role.

• As was mentioned above, the scope of the data scientist role can vary broadly, and this scope isn't always obvious from looking at job titles. Identifying the specific type of work that you need your new hire to tackle is the first step to deciding what skills you need to look for on their resume.  Some of the common responsibilities of a data scientist include: 

  • Mining or extracting usable data from a variety of sources
  • Processing structured and unstructured data to prepare it for analysis
  • Cleaning and validating the integrity of the collected data
  • Analyzing large quantities of data to identify patterns and trends
  • Designing prediction systems and machine learning algorithms that can solve specific business problems or predict future trends
  • Presenting and reporting on the findings from data analysis
  • Creating visualizations and models that aid understanding of the data
  • Developing solutions to business problems based on data analysis
  • Collaborating with other teams, like marketing, IT, product development, or leadership, to determine the requirements and goals of data analysis
  • Ensuring that data collection and handling complies with legal regulations and the ethical standards of the organization
  A data scientist's day-to-day responsibilities could include all of the items on this list, or they may be asked to provide more niche services in one particular area. 

• Step 2: Craft an effective job description.

• In theory, writing job descriptions is fairly simple and straightforward. There are three key sections that you need to include: a description of your company and the job; the responsibilities of the role; and the qualifications, experience, or skills that you're looking for in a candidate.  The first step to writing an effective job description is understanding the position well enough to accurately describe it. Having discussions with team members who work similar roles currently can be a big help here, especially in a technical field like data science. Find out what specific tasks they handle most on an everyday basis, as well as what skills are the most valuable or necessary for them and what past experiences best prepared them for their current role. You can also look at job posts for similar roles from other companies to see the ways they outline the role's responsibilities and the requirements they highlight as key indicators of a candidate's likely success.  Don't neglect the company description part of the job posting, either. In a competitive field like data science, the job description needs to do more than just explain the opportunity to interested applicants. It also needs to convince them why they should work for your company, and your mission, values, culture, and work environment are often a major factor in those decisions.  Another way you can entice potential applicants to your company is to highlight the benefits your workplace offers employees. This doesn't just mean the salary, health insurance, and other compensation you offer, although that is a part of it. Also consider how working for your company could benefit a candidate's work/life balance or career progress. If you offer hybrid or remote options, a generous PTO policy, or other types of workplace flexibility, that is definitely something to highlight. The career advancement side can include things like professional development, upskilling, mentorship, or internal mobility. Finally, make sure that your job postings are written in a way that encourages all qualified candidates to apply. Review your job descriptions for gendered or biased language that could make individuals from underrepresented groups feel excluded or unwelcome. In a field like data science, where qualified talent is often in short supply, staying open to candidates from all backgrounds is especially important. 

• Step 3: Source candidates.

• Once you've perfected your job description, it's time to share your opening with the candidates who could be a good fit for it. The first place that many employers do this is using online job board platforms. There are a variety of these websites to choose from. In addition to general employment marketplaces like LinkedIn, Indeed, and Glassdoor, you can focus your search by using niche job boards like: 

  • AI-Jobs.net - An international platform covering a wide range of careers in AI/ML, big data, and data science. 
  • Built in - An online community focused on tech companies and startups, they have one of the largest job boards for the data niche.
  • DataJobs - This board is focused on helping employers connect with data analytics talent from around the world. 
  While these platforms can be useful, they share an inherent weakness: they're only likely to connect you with people who are actively looking for a job. The high competition in data science often means companies can't afford to wait for the perfect candidate to stumble across their posting. Instead, they need to be more proactive in seeking out talent pools. You can do this in a number of ways, including:
  • Attending job fairs, conferences, and other industry events
  • Connecting with professional organizations like the Association of Data Scientists (ADASci), the Data Science Association, or the International Institute of Business Analysis (IIBA)
  • Partnering with local universities to offer internships or entry-level roles to students and recent graduates
  • Reaching out to qualified professionals you find on LinkedIn or in online professional communities
  • Hiring a recruiting firm that specializes in data science or technology

• Step 4: Assess candidates' technical skills and experience.

• Regardless of your industry, it is rare to get a candidate who possesses every single skill, trait, and experience listed in your job description. This is especially true in a field like data science, where the technology, tools, and required skills are constantly evolving and advancing. Before you start your candidate assessments, it's important to identify which skills are must-haves for a candidate and which are things that would be nice to find but can be trained or added after hiring. The specific data science skills that are most necessary will vary depending on the role, but they often include: 

  • Proficiency in programming languages, including Python and R
  • Familiarity with libraries like NumPy, pandas, and matplotlib
  • Experience with cleaning, organizing, and manipulating data
  • Thorough knowledge of statistics, probability, calculus, and linear algebra
  • Proficiency with SQL and relational databases
  • Familiarity with NoSQL databases like MongoDB and Cassandra
  • Experience with data visualization tools like Tableau and Power BI
  • Knowledge of common machine learning and deep learning algorithms and techniques
  • Experience with cloud computing tools like Amazon Web Services (AWS), Microsoft Azure, and Google Cloud
  • Knowledge of big data tools and techniques, such as database management, ETL processes, and real-time data analysis
  The first step most employers take to assess whether candidates have these skills is to look at the education, certifications, and previous work experience on their resume. You can also give applicants test assignments or have them take targeted skill assessments to verify their capability in key areas, or gauge the depth of their knowledge by asking targeted questions during interviews. 

• Step 5: Evaluate soft skills and cultural fit. 

• While the technical skills needed often get more attention when you're hiring into technology roles, soft skills and cultural alignment can be an even better indicator of the candidate's long-term performance in the role.  The most valuable soft skills for data professionals include: 

  • Communication - Just because someone works with data doesn't mean they never interact with people. Data scientists need the ability to share the results of their analysis with company leaders, coworkers, and other stakeholders who lack a technical background. Communication skills like active listening and empathy also enable them to better understand the goals of the analysis, and to better interpret customer behavior and feedback in their analytic efforts.
  • Business acumen - Understanding the industry and broader context of the data to be analyzed helps data scientists to conduct more targeted analyses, yielding insights that are more consistently useful for addressing business challenges or problems. 
  • Critical thinking - Data analysis is often used to solve complex problems. Before they even begin their analysis, data scientists need to make sure they're asking the right questions and focusing on the right of information. Having strong deductive reasoning, problem-solving, and other critical thinking skills allows them to do so. 
  • Adaptability - As technology advances and the data landscape evolves, professionals who work within it need to be prepared to pivot their skills and practices to match. A professional who is flexible and adaptable will be better able to make these adjustments, keeping their skills current with the business' needs. 
  The best time to assess a candidate's soft skills is during the interview phase. You can use situational and hypothetical interview questions to gauge how the candidate approaches their job.  Calling applicants' references can be another way to identify a job seeker's soft skills, as well as whether they're aligned to your culture. You can ask questions about how the candidate functions within a team, how well they communicate, and other aspects of their personality and behavior that can help you decide if they'll be a good fit for your team. 

• Step 6: Make your final hiring decision and extend an offer.

• Once you've reviewed all of the applicants and determined whose skills are the best match for your needs, it's time to extend your offer. Ideally, you want to do this as soon as possible after posting the opening. Remember, there's a strong chance that your top candidates are being courted by other employers, too. The sooner you can send them your offer, the higher your odds of adding them to your team.  Be prepared for some negotiation to happen during this stage, as well. Your initial offer should be on the lower end of what you're prepared to pay, allowing you some room to adjust upward in response to counter-offers from candidates. Keep in mind that you don't need to wait until the offer stage to bring up things like salary expectations, starting dates, and other details that are included in job offers. Discussing these details during the interview, or even including this information in your job posting, can help to streamline the offer negotiation process and ensure you and the candidate are on the same page.

• Hiring the right data talent for your business

• Finding and hiring data professionals can be a challenge in the current employment market. Taking the time to fully understand your requirements for the role can streamline your search, but it often still requires a more proactive approach to attract the right applicants. Following the steps and tips outlined in this article can help you to hire the right data scientist, right when you need them.