It’s been three and half years since the Harvard Business Review declared that data scientist was the sexiest job of the 21st century. Since then the need for data scientists has skyrocketed but the supply has not, which puts recruiters in the delightful position of having to compete with more and more recruiters for a dwindling supply of talent.
Without a strong funnel of data scientists graduating from colleges or grad programs, it’s unlikely that the shortage will end any time soon, which means we’ve gotta hack our way out of the data scientist cul de sac.
Here’s a few tips to help you find that proverbial needle in the haystack.
Carefully consider your data needs first.
Not all data jobs are the same, which means you may not actually need to hire a data scientist. The benefit of not hiring a data scientist, and say, trying to hire a data analyst, is that it could save you a ton of time and money. According to Payscale, the median salary of a data analyst is $55,000 while the median salary of a data scientist is $93,000. Just by listing a job title as Data Scientist rather than data analyst, you’re almost doubling the compensation expectation and putting yourself in competition with huge companies that can probably afford to pay them a lot more than you can. As one analyst told Fortune Magazine, “the very term pushes their value up.”
You may not need a full-blown data scientist with a PhD and eight years of experience. You may just need a data analyst or a data engineer who can help you create better data systems or clean your data, neither of which necessarily require a data scientist’s advanced skillset. Carefully calibrating your search based on your actual data needs will save you and your organization precious time and money, and allow you to focus on the precise skills and qualifications that are most relevant to you.
Train your current employees in data.
Since it’s so hard to hire data scientists, a better move is to turn your existing employees into data nerds! These days there’s a wealth of options for learning data science, including online courses, classes, grad programs and boot camps. Given the expense and difficulty with finding data scientists, facilitating the training of interested employees in the fundamentals of data science is a great alternative method for building out your data team. If you don’t already have a continuous learning perk available to employees, let this be the test program.
While being a great data scientist requires a mix of hacking ability, math and statistics knowhow, and business savvy, people can learn to analyze, visualize, and communicate data findings. Udacity, Udemy, and Coursera are just a few resources to start browsing online data science courses. One of the most important factors for many employees, especially millennials, is growth potential. What better way is there to help ensure you keep your talented technical people and add certain high need skills?
“Crowdsource” data scientists and test applicants on very specific skills.
If you’re struggling to build a talent pipeline of qualified data scientists, give them the ability to self-select and test into your application process. Post a “take-home” test on your website for interested candidates, and design it to test basic analytical and technical skills that are the minimum a person needs to contribute as part of your data team. Start with the organization’s most pressing problems on the data front and design a test for those problems. Give a set amount of time to submit their completed test (like 24 hours or three days) and then bring anyone who passes in for an interview (you can grade on a curve too). Make sure your test resembles the real data problems your organization is facing – that will give you a far better idea of how they’ll fare as part of your team.
By lowering artificial barriers to entry and making the process highly specific to your company, you’ll drive more people into your funnel and discover hidden diamonds you may have missed otherwise through more traditional sourcing. They may not be perfect, but if they are skilled, trainable and have the building blocks, they’ll become valuable assets as you build out your data team.
Hire efficiently (...and we mean it).
In a hiring market as competitive as data science, it’s a sprint to the finish line of signed offer sheets. You need a streamlined, well-defined process that cuts out unnecessary bureaucracy. Here’s how to do it. Map out your process step by step (here’s a great explanation of how Sailthru, a data analysis company structures their data hiring). For instance, after the “take-home” test, have a tech savvy team member perform the initial phone screen to be sure your candidates can talk the talk. If they pass the phone screen, get candidates in as soon as possible for in-person interviews and hands-on work with the team and hiring manager. If every stakeholder approves, be prepared to make a prompt, yet well-calculated offer.
The idea of hiring without endless checks and balances can be difficult to implement, since it feels inherently risky. This is where a tried and true structure for your entire hiring process pays dividends. As tempting as it is to look for perfect hires, that leads to false negatives, a costly point of failure where you don’t hire people who in reality would make amazing additions. By standardizing the process, you can avoid getting bogged down in unnecessary debates, disagreements or ambiguity that costs you time and hires. When it comes to hiring data experts, you just can’t afford it.
The Best Tech Literacy Tools for Non-Technical Recruiters
How to Evaluate Technical Candidates as a Non-Technical Recruiter
Entelo Data Shows the Best Time for Recruiting Salespeople