The first step in hiring the right data scientist is being sure your organization is ready for one. While companies talk about “big data” and “analytics” almost as if they were silver bullets, the field is still new enough that some underuse their analytics staff by having them focus on compiling reports rather than putting their skills to use in a more strategic, meaningful way.
When the time comes to hire a data scientist, your focus should be on selecting the right data scientist – the one who’ll bring the most value to your company’s particular challenges. Here are five ways to make sure you do just that.
- Have a clear understanding of what you want the data scientist to do. Don’t create a blue-sky job description. Instead, think of several challenges you need the data scientist to address, and be sure you have the necessary data available or ready to compile. Then use this material to develop problems the candidate can work on during a day in your office. Be sure your team is available to answer questions and pitch in a bit. At the end of the day, have the candidate present her solutions and describe the methodology she used to get there.
- Include your team in the hiring decision. If you give your team the chance to interact with the candidate, they’ll make valuable observations about how he works, thinks and approaches problems. Plus, they’ll have a sense of how well he’ll fit into your department’s and company’s culture. While data scientists need to have the right quantitative and technical skills, cultural fit counts when it comes to ensuring you make the right hire.
- Get opinions from outside the department, too. To succeed, the candidate must be able to communicate her ideas and methodologies effectively, and to take complex concepts and make them understandable. Make sure she can bridge that gap before making your decision. To make real contributions to the company, the data scientist has to know your business.
- Probe for business knowledge. To make real contributions to the company, the data scientist has to know your business. So probe about his sector expertise, and ask other departments to do the same. Does the candidate think in ways that allow him to recognize opportunities that others might miss? Does he have a streak of creativity that prompts him to look at the business in new ways? Here’s where the input from product, marketing and sales people can be particularly important.
- Be clear on the kind of data they’ll be working with. In the Harvard Business Review, data scientist Michael Li contends that many hiring efforts fail because the employer isn’t clear on whether they need analytics for humans or machines. Those who produce analysis for computers, Li said, require strong mathematical, statistical and programming skills so they can build dependable predictive models. Those who produce results for people must be comfortable communicating and working at higher levels, focusing on the “how” and “why” of their conclusions.
Hiring a data scientist isn’t like hiring for other roles. The mix of quantitative, technical and communications skills required can very greatly from position to position. Approach the process thoughtfully and have a clear idea of what you need this hire to accomplish, and you’ll greatly increase your chances of landing the right candidate.