Q Ethan McCallum, a strategic consultant in data analysis and technology, describes the Common Mistakes in Data Science Hiring. Is he describing your hiring efforts?
Doing Data Science Hiring for the Sake of Saying You’re Doing Data Science
Figure out what you want to do and why. Having a purpose and an expected ROI are critical questions to ask. Jumping on the latest “cool” bandwagon for the sake of coolness doesn’t work.
Connect the what you want to do with how will you do it. Map out your data sources and the skills needed to do the job.
Diving into Advanced Data Science as a First Step
You must figure out how to clean your data, store your data and the technology infrastructure needed before you can even do the first stage of data science.
Having HR be the First Contact with Data Science Candidates
This very competitive marketplace means hiring managers have to do their own outreach to convince great data scientists to talk with your company. General interview questions will not make your position desirable and exciting.
Waiting on Perfection
Finding someone who can clean the data, code, create software, do analysis, interpret and present results all rolled up into one person is a tough order. Be realistic in your job requirements.
Unrealistic Interview Techniques or Metrics
Limit technical questions to tools you actually use and the ways you use them. For experienced professionals ask open-ended questions. If you insist on a presentation or white board session, make sure it is close to what you will have them do on the job.
Making the Data Scientist the First Data Hire
You must have a solid data infrastructure and data engineers to prepare the data for the data scientist to use. Get those roles in place first.
Hiring a Junior Level Data Scientist First
Yes, the salary required for an experienced data scientist is significantly higher than someone right out of boot camp. While it is easy to quantify the savings on salary, the hire won’t be productive. They will have no one to turn to including themselves.
Insisting on DIY
Hiring the best person you can and just throwing them in to your problems isn’t the best strategy. Even for those candidates who can make it work and don’t leave in frustration, it will take them much longer to be productive.
Demand Instant Results
Be patient. The work data scientists do can be incredibly impactful on your business. Give them a chance to do their best work.