Data Science Interviews – Why Are You Falling at The Last Hurdle?
22 October 2018
Interviewing for a new job is a full-time job in itself. With the average interview process taking 45 days for a Research Scientist/Data Scientist - Glassdoor, it takes a considerable amount of commitment in time, preparation and energy. Even more so, that commitment can be a whole lot of wasted time, preparation and energy if you fail to receive an offer because you did not live up to the expectations of the hiring manager. As a headhunter in the Data Science space, I see candidates with amazing backgrounds and credentials fail on the same key areas during an interview, and believe me it is not because the hiring managers do not want to hire.
Embellishing your resume (Cue the buzzwords)!
Resumes or nowadays LinkedIn profiles are a tool that provides an overview of your achievements and experiences. So why lie or “stretch the truth” when this is the only thing that an interviewer has as a source of information on yourself. Listing that you built Machine Learning models when you have in fact not, will only show you out to be a liar, and put all your other experiences and achievements (which you worked hard for) in to question. A personal favourite which is becoming more common nowadays is the over exaggeration on experience with Neural Networks. Filling your resume up with buzzwords such as Machine Learning, Neural Networks or Reinforcement Learning will probably get you an interview, but when you can’t prove that you have done what you said, there will be no job offer.
I suppose the moral of this story is, be truthful with yourself and the interviewer, as you will almost always be caught out (just saying).
Did you know 7/10 Data Scientists fall short on statistics – and that’s a stat (fact)!
Who would have known that statistics is a core competency of Data Science? As we all know that it is, why do so many fall short on interview questions that involve areas such as core statistical concepts, probability, Bayesian thinking, and statistical machine learning? Well, the answer is – they do not prepare for these questions! After speaking to hundreds of Data Scientists the common theme that arises is summed up perfectly in this comment “I didn’t practice or prepare around statistics because I thought that was the easy part of the job”. This is true guys; Data Scientists fail on statistics because they do not prepare for these questions. YOU WILL GET QUIZZED ON STATISTICS!!
The moral of this story can be described in 6 P’s; Proper Preparation Prevents Piss Poor Performance!
Lost in translation!
So, you’re an Ivy League graduate, maybe with a PhD. You are universally seen as one of the smartest people on the planet, BUT you can’t provide a non-technical explanation, or explain what your work has done to non-technical stakeholders. Data Scientists regularly provide analysis to non-technical colleagues across different areas of an organisation from Sales to HR, so it is paramount that you have the necessary skills to deliver the analysis to its full effect. Many times I have received feedback from a hiring manager that echoes the above statement. You can be the most technical and smartest Data Scientist in the world, but what purpose do you serve to an organisation if you can not convey your findings which in turn will provide business value.
Moral of this story, remember who your audience is and remember the purpose of your work.
Anyway, I’m a headhunter that is enough morals for me today (this is a joke). Data Science is one of the most rewarding careers there is. On a daily basis, you get to empower important business decisions and actively be involved in change so don’t go and fall at the last hurdle!
If you have any questions or are a data scientist looking for your next role, then please connect with me on Linkedin and/or get in touch - firstname.lastname@example.org