Hiring the wrong candidate can be an expensive affair. Not just in terms of the onboarding expenses, but long-term repercussions such as project delays, below average results, increased attrition rates and others. While the companies take a lot of precautions and elaborate procedures such as a combination of technical interviews and coding assessments along with soft skills to make sure they are hiring a perfect candidate, there are chances of them going wrong. More often than not, there are other biased factors that might lead to a wrong hire, such as a candidate having a vibrant personality, candidates having come through referral, among others.
It is therefore important to make a sound decision while hiring a candidate and overcome the challenges of having a wrong hire. In this article, we talk about five such measures that companies can take to reduce the chances of hiring a wrong candidate in data science.
1| Not falling blindly for keywords such as machine learning, NLP, AI and others: Most of the times, the job requirements in data science end up mentioning key skills such as AI, NLP and others as must-required in a candidate, which are conveniently mentioned by most of the candidates in their resumes even if they are not completely acquainted with the terms. Using application tracking that only picks up resumes with these keywords may, therefore, prove detrimental. It is important for recruiters to analyse the genuine interest and knowledge of the potential candidate in these areas to stop from adding just another headcount in your team.
2| Conducting hands-on tests and hackathons: In an interaction with Analytics India Magazine, Srinidhi Rao of TheMathCompany had said that candidates recruited through hackathons have a lower chance of being wrong hires and we couldn’t agree more. Tests and practices where actual analytical skills and understanding of the tools of a candidate are exposed can only prove to be quite helpful. Hackathons are one of the tried-and-tested solutions to bridge the gap between theoretical and practical knowledge. As opposed to traditional methods of candidate evaluation, hackathons simulate a real-world environment for candidates with components like solving real business problems to working in a deadline-based environment.
3| Overlapping of designation such as data science, analysts etc: It also happens quite often that companies end up posting out a job requirement that is quite different from actual challenging and thought-provoking work that data scientist might be looking for. For instance, many companies use the words data science, data analysts, business analysts and many other designations interchangeably, that might confuse the candidate, often ending up being a wrong hire. It is therefore important for companies to put out very specific job requirements specifying skills in data science that they are actually looking for.
4| Ensuring that the candidate’s and company’s interests are on the same page: While a company may find a candidate with all the required skills that they want, it is necessary to cross-check with the potential candidate on the kind of opportunity and growth that they are looking for. It might end up being a disaster hire if the end goals of employees and companies do not lie on the same page.
5| Do not rush into hiring the first data science candidate you interview: There is a huge demand for data scientists today and companies are in a rush to hire as soon they get one. Rushed hiring may end up into being a wrong hire as there may be chances of not getting the background check right. In a race to be data scientists, there are a lot of people who like to call themselves data scientists just based on Excel or some basic visualisation tool knowledge. This might later turn out to be disappointing as a data science job role is much more than that.
