Big data means big ideas, and for recruiters, it means even bigger duties.
Success in hiring good data scientists requires a keen eye for spotting talent with a vast, highly specialized and applicable skillset. As with the buzzphrase “big data,” the data scientist position is quite nebulous and spans across a myriad of strengths in areas including math, data engineering, analytics, and computer science.
SmartData Collective even called the data scientist position the “sexiest job no one has” — sexy because of its versatility and relevance, although the talent gap is arguably unattractive. Despite the job's appeal, only about a third of candidates are qualified to fill the role, and by 2018, the U.S. could experience a shortage of nearly 200,000 data scientists.
“...Data scientists are the designers and content creators of today, not the software engineers or the IT bottleneck.” — Omer Trajman, Partner, Scaling Data
In fact, run a search for data scientists online and it’s likely you’ll find there are very few results that pop up. For one, the position is new and emerging — at least under 10 years old. Also, the number of candidates who are data scientists and mark themselves as such are few and far between either because they’re 1) already working for a top organization or simply because, 2) many other individuals who perform data scientist-esque work are really other kinds of engineers or applied scientists wearing other hats for the sake of their company’s limited resources. Doing more with less, so to speak.
The hybrid nature of the data scientist merges programming, analytics, and the ability to visualize and communicate how a company can transform information into items like a product feature, actionable intelligence, or a new approach to marketing.
“A data scientist does not simply collect and report on data, but also looks at it from many angles, determines what it means, then recommends ways to apply the data.” — IBM
The key to hiring data scientists? Skip the interview questions and pop quizzes, and let the candidate work on your company’s challenges in real time.
Hire for a data scientist who is able to adapt to rising trends and models with great fluidity and adept shrewdness. One key role of data scientists is to be able to pick and analyze relevant data in order to predict trends and help a company adjust their business strategies accordingly.
“A good data scientist is someone who has the right tools (math, programming, critical thinking), is self-sufficient (doesn't need someone else to implement his or her ideas) and has an interest in understanding the context in which the skills can be applied. This is what the marketplace seeks.” — Mark Bregman, Senior Vice President and Chief Technology Officer, Neustar
Even Riley Newman, AirBNB’s head of analytics and data science, swears by this tactic: Allow data scientist candidates to experience a day in the life at your organization, working in tandem with fellow product, developer, and engineer team members to complete projects and decipher a real challenge the company hopes to solve.
For the company and candidate, it’s a win-win situation, regardless of whether or not the candidate is actually hired. Companies are given the chance to show off their teams’ smarts — their capabilities and creativity in developing a product. Candidates are offered a channel for demonstrating their own skills, provide valuable input and ideas, and practice for future rounds of tough coding, reiterating, and evolving.
Allowing candidates to demonstrate how they perform on the job offers companies an authentic peek into their process for solving real issues at hand and how they interact with employees who may become their future coworkers. The contract-to-hire onboarding process is ideal, as it enables hiring managers to evaluate whether or not a candidate is a good fit over time, but the solution is likely viable for the short-term, and going through a consequent hiring process expends unnecessary time and costs.
Hire once, hire right, hire a data scientist for life. Not really, but you get the idea.
