PRACTICAL GUIDE
What I've learned from interviewing more than 300 Data Scientists
Preparing for an interview to stand out and increase your chances of getting the desired job.
I never expected to write an article on interviewing Data Scientists since there have been so many blogs and articles on the subject. Don’t believe me? Just google terms like ‘how to best prepare for an interview,’ and you will be flooded with results.
So, what prompted this article? As often, it stemmed from a conversation. In a catch-up call before the holidays, my friend Tom told me that his daughter, a Data Scientist, was having difficulty finding a decent job since she wanted to switch businesses. That took me by surprise. I know her quite well, including her strong technical and academic background. I wondered if I could be more specific in talking about my experience recruiting Data Scientists and explain why certain applicants thrive while others fail.
I sat down to reflect on the two years I spent building up the Mars Petcare Data Science team, reviewing my entire interview notes from entry to senior level. Not surprisingly, some clear patterns emerged from the 300+ interviews, and below are vital commonalities that separated candidates from succeeding and those who ended up not progressing in the interview process.
Communication and Storytelling
Communication is a big part of a Data Scientist’s job — whether it is technical communication with peers or translating technical concepts to business audiences and stakeholders. The more senior the role, the more influencing you need to do, which all comes down to your storytelling capability. You may be the strongest technical candidate, but you need to clearly articulate why you made certain decisions, identify caveats and limitations, and be concise in your answers, especially when you talk about the impact and route to the value of the work completed. The interview panel eliminated most candidates due to poor or inadequate communication for the level they were interviewing for (from entry-level Graduate Data Scientist to the most senior role of a Principal Data Scientist).
Recommendations to solve this potential pitfall: in preparation for the interview, write down the context of the things you highlighted on your resume, why you picked this solution over another, and most importantly, what the impact or outcome was, including whether or not the business implemented your solution. The more you practice with technical and non-technical audiences, the easier it will get. An additional tip for your practice rounds is to change the level of detail, going from executive and high-level summaries to detailed technical deep dives.
Second, take advantage of resources like Toastmasters International, where you can practice public speaking and improve your communication skills or dive into the great content published by the likes of TED Talks. I’m not affiliated with or sponsored by either organization.
Company and Business Understanding
As Data Scientist, your job is to create real solutions that help advance the business. This means either enhancing or optimizing a process, identifying new opportunities for growth and innovation, or more altruistically leaving the world a better place (e.g., as Mars Petcare, one of our objectives is to create a better world for pets). Since these are your objectives as Data Scientist, it is critical to understand the fundamentals of the company you are interviewing for. In addition to not being able to connect with the interviewing panel about the problems that matter the most to them, how will you know if this position is the right fit for you as you progress in your career? How will you be able to ask meaningful questions and engage with the interviewer? Also, candidates who didn’t have any questions to ask at the end of an interview, not a single one, ended up not making it to the offer stage. I’m not saying this was the reason. Still, it appears like a strong indicator of having some basic understanding and curiosity about the business, its challenges, and how you, as a Data Scientist, are expected to support it.
Recommendations to solve this potential pitfall: to stand out, you need to invest some time to understand the company you are interviewing for, including its primary purpose and positioning in the market, source of revenue, and significant competitors. As you advance in your interview process, continue to dedicate more time to discovering the company. This investment will allow you to better connect with the interviewer and enable you to understand if this is a suitable workplace. Regarding resources, start with the corporate website to get a feel for the brand and what it stands for. Look for some financial information through the investor relations section or other publically available resources, including Wikipedia or Yahoo Finance. Seeing the environment where the product or service is being sold could also lead to some significant observations and talking points during the interview — for instance, going to a retail store, or browsing the eCommerce website. Lastly, your Linkedin network could be a more utilized resource. Reach out to your first and second-degree contacts, convey your interest in the firm, and ask for 15 minutes of their time to learn more. Yes, you will be rejected or not replied to, but in the end, you only need one or two chats to distinguish yourself from other candidates.
Two more things that kept getting high marks from the interview panel
- Demonstrating you have a keen sense for identifying pragmatic solutions and a strong product sense: this would, for instance, come through you demonstrating a strong focus on the impact and outcome of the work rather than the work itself. I talk about the importance of product sense here, in case you want to learn more.
- Be your authentic self, and don’t assume anything in the conversation: ask clarifying questions when needed, as this will provide a good insight into how you will engage and speak with senior and executive stakeholders.
Conclusion
Where we work and what we do most of our day are so important. Your preparation will set you apart from other candidates, especially in how you communicate, tell your story, and demonstrate your grasp of the company. Ultimately, the hiring manager will decide whether he trusts you to do the following — the more senior, the higher the required trust.
As the hiring manager, I ask myself whether I can trust the candidate to:
- complete an assignment
- solve a problem
- work well with others to solve a problem
- figure out which are the correct problems to solve
- inspire other people to solve the correct problems
As always, curious about your experiences and thoughts. Please let me know and leave a comment.
