8 min read
One of the biggest factors that helped me save $100,000 by age 24 was landing a job as a data scientist with an entry level salary of $80,000. Earning this high income in a low cost of living area (Cincinnati) has allowed me to save thousands of dollars each month, which I have been able to funnel into investments and savings accounts.
Since sharing about how I became a data scientist, I’ve received several emails from readers asking if I could provide specific details on the actual interview process, including:
How did I first get in contact with the company I currently work for? Did I submit a resume on a massive job search site like Indeed? Did I know someone within the company? Did I receive a message from a headhunter on LinkedIn?
What was the interview process like? How many rounds of interviews were there? Were the interviews one-on-one or were they in a group setting?
What specific questions did I have to answer in the interviews? What type of technical questions did they ask? What type of non-technical questions did they ask?
In this post, I answer all of these questions in detail.
Even if you don’t plan on becoming a data scientist, you may find this post interesting. I’m always interested in what the interview process is like for people in other professions like investment banking, engineering, teaching, graphic design, etc. It’s neat to peer behind the curtains of the interview process for different professions.
Before I answer specific questions about the interview process, I should mention the following background details about my professional experience during the time when I was searching for a data scientist job:
- I had a bachelor’s degree in statistics and a master’s degree in applied statistics.
- I had one year of experience as a data analyst intern and one year of experience as a predictive data analyst at a healthcare company.
- On my resume I listed that I was proficient in the Microsoft Suite, the SAS programming language, and SQL.
These qualifications played an important role in getting my resume noticed in the first place.
How did I first get in contact with the company I currently work for?
I received zero responses from the companies on Indeed, one response of “you don’t have enough experience” from a company on Monster, and zero responses from the companies on CareerBuilder.
Luckily, I received a message out of the blue from a headhunter on LinkedIn asking if I would like to chat about a potential data scientist position.
Headhunter: a person who identifies and approaches suitable candidates employed elsewhere to fill business positions.
This headhunter found me on LinkedIn because I had configured my profile settings to show that I was “open to new opportunities.”
Check out this short guide on how to configure your own LinkedIn settings to let recruiters and headhunters know you’re open to new job opportunities.
I said “yes” and just a few hours later I jumped on a ten-minute call with her. During the call, I answered the following questions:
“Do you have experience with either R or Python?”
Me: Yes, I have been using R for two years in a professional setting.
Note: Although our primary programming language at my day job at the time was SAS, I had been learning R in my free time and had even convinced management to let me use it for a couple projects.
“Do you have experience with SQL?”
“Do you have experience with presenting findings from an analysis to upper-level management?”
“What is your desired salary?”
Me: I’m looking for a minimum of $75,000.
Note: At the time, I was earning $52,000. Throwing out the number $75,000 took some courage. It turns out I should have asked for even more, though.
Once this interview ended, I emailed the headhunter my resume and she said she would follow up with me within one week.
Fast forward one week. The headhunter spoke with the company. The company decided they wanted to move on to the next stage with me, which involved completing a coding challenge in R where I had to solve four problems in one hour.
I completed and submitted the coding challenge. Although I never actually received my results to know if I answered the questions correctly or not, I received a call a couple days later from the headhunter letting me know that the company wanted to proceed to the next stage, which included two in-person interviews at the company’s building.
What was the interview process like?
The interview process was comprised of two one-hour interviews with two members of the data science team. The interviews took place one week apart from each other.
Before the interviews, the headhunter let me know that the company had a casual dress policy, meaning the employees were allowed to wear jeans and a t-shirt, as long as the t-shirt had the company logo on it. Despite knowing this, I wore a suit just so the company knew I was serious about the opportunity.
Each interview took place in a fairly small room with one table in the center and a whiteboard on the wall, which I ended up using to solve various questions they asked me.
In general, each interview was very laid back. It was nice only talking to two members of the team at once, as opposed to a group of four or five.
What specific questions did I have to answer in the interviews?
I answered both technical and non-technical questions in both interviews.
For the first interview, I met with one manager and one team lead. Here’s how it went:
Them: “So tell us a little bit about yourself.”
Me: “Well, I grew up in _____ and went to school at _______ where I obtained my bachelor’s degree in statistics and my master’s degree in applied statistics. During my time at ______ I was a research manager in the statistics department where I consulted with other students working on projects that involved statistical analyses. Once I finished my bachelor’s degree I worked at ______ as a data analyst intern and later as a predictive data analyst on their data science team.”
Them: “Tell us about your favorite project you worked on while you were at _______.”
Me: “My favorite project involved working with the marketing department where we had to use k-medoids cluster analysis to segment the consumer market that our company targeted into seven or eight distinct groups based on their socioeconomic status, spending behaviors, and health risk. This project involved gathering the raw data with SQL, cleaning and munging the data with SAS, building a model with R, and eventually passing off the final model to the development team which implemented the model using Java into a production environment.”
Them:“Why are you interested in joining our team specifically here at _____?”
Me: “I’ve read online about several of the projects the company is working on within the data science team here and I’m highly interested in working with a team that does such innovative work that has such a big impact on the company as a whole. I also think that based on my educational background and my experience that I could bring a lot of value to this team.”
Them: “What is your favorite R package?”
Me: “Probably ggplot2 because I use it in almost every analysis either for exploratory work to better understand the data or for visualizing the final results of an analysis.”
Them: “How do you typically get data into R to work with it?”
Me: “I either read in raw CSV files that contain data or I query directly from a database using SQL within R with the RODBC library.”
Them: “Two numbers multiply to 180 and add up to 27. What are the two numbers?”
Me: 15 and 12.
Note: I was allowed to use the whiteboard for this question, but I figured it would be more impressive if I could just think through it in my head. It took about 10 or 15 seconds for me to come up with the answer.
Them: “What is the definition of correlation?”
Me: A metric that measures the linear relationship between two variables.
Them: “Can two variables be related to each other while having a correlation of zero?”
Me: No, I don’t believe so.
Note: I got this question wrong. Two variables can have a non-linear relationship (like a quadratic relationship) while having a linear correlation of zero.
For the second interview, I met with the director of the data science department and one manager. Here’s how it went:
Them: “So, what made you interested in data science?”
Me: “I’ve enjoyed working with numbers ever since I was a kid. I find them fascinating to work with and math has always been easy for me to understand. This naturally lead to me studying statistics in both high school and college, which eventually lead me to become interested in working with data to answer interesting questions and solve hard problems. From there, I began to read data science books and take online courses and grow my knowledge base. I find it to be a fascinating field and also a useful one.”
Them: “Tell us about your least favorite job you’ve ever had.”
Me: “Probably when I was a cashier at a grocery store in high school because I felt like my job could be automated fairly easily and that I wasn’t really making a difference in the company or growing my skill set in any way.”
Them: “Tell us about a time when you went above and beyond what you were asked to do at a job.”
I have always found this question to be cheesy to answer. I ended up giving some answer about how I helped a customer walk through an entire grocery store when I used to work at one in high school to help them complete their shopping list because their vision was partially impaired.
Them: “How many ping pong balls can fit on a school bus?”
For this question, the actual number I provided didn’t matter. They were simply interested in how I would go about solving the problem. So, I asked them questions like, “What are the dimensions of the bus?”, “What are the dimensions of a ping pong ball?” and “Is the bus completely empty?” and based on their answers, I did some quick multiplication to estimate how many ping pong balls could fit on the bus.
Them: “You know those manholes in streets that lead to sewer systems below cities? Why are those manholes in the shape of circles rather than squares?
For this question, I had no clue. I ended up giving some answer about how circular manholes are cheaper to produce in a factory so they’re more cost-effective to produce at scale.
It turns out that the correct answer was that a circular manhole, when removed from the street, can’t fall through the hole. By contrast, a square-shaped manhole could accidentally fall through the hole if it was turned diagonally, since the diagonal of a square is longer than the side of a square.
Them: “Two trains are 150 miles apart traveling towards each other. One is moving at 18 miles per hour and the other is moving at 12 miles per. How many hours until they converge?”
The answer is 5 hours. It took me about 3-4 minutes to solve this one on the whiteboard and I actually got the answer wrong twice before I was able to correct myself. The way to solve this problem is through using linear equations. Check out this page for an example of how to solve a problem similar to this.
Them: “Give us an example of a variable that is normally distributed.”
Me: The height of everyone in this building.
Note: There are many variables that exist in the world that are normally distributed. Height and weight are the two most common.
Landing the Job
After completing both of the interviews, the headhunter reached out to me about a week later and let me know that I had landed the job.
Since this was a contract-to-hire position, the headhunter and her agency actually did the salary negotiations for me. They ended up negotiating an hourly rate that was equivalent to a $80,000 salary. Since this was more than my minimum of $75,000, I accepted the position and started about three weeks later.
Crafting the Perfect Resume
The one thing that got my foot in the door at this company was having a resume that caught the attention of a headhunter.
I’ve written before that being able to craft a good resume is often a better way of landing a job than going back to school and getting additional degrees or credentials.
For those who are interested in getting into the field of data science, it’s helpful to see real example of resumes that have landed data scientist interviews.
Fortunately, I have three real examples of resumes that did successfully land interviews and eventually three job offers as well: my own resumes!
I have compiled a package of my exact resumes that helped me land three data science jobs, all at different levels.
The first resume helped me land my first data science internship as a senior in college with no full-time job experience. This job paid $15 per hour.
The second resume helped me land my first full-time job as an entry level data analyst at the same company. This job came with a salary of $52k.
And the third resume is the one that helped me land my current job as a data scientist. This came with a salary of $80k.
This package of resumes offers serious value to aspiring data scientists because I’m not just giving you generic tips on how to improve your resume; I’m showing you the exact resumes that help me land real jobs. This package is a resource that I wish I had access to three years ago when I was clueless on what employers were actually looking for in a candidate.
If you’re an aspiring data scientist looking for examples of actual resumes that have landed data scientist interviews, click on the button below to download this package of resumes for $15:
- Investing is About Absolute Numbers, Not Percentages - September 23, 2020
- Sunday is for Sharing: Volume 170 - September 20, 2020
- How to Apply Zipf’s Law to Your Finances - September 16, 2020
Full Disclosure: Nothing on this site should ever be considered to be advice, research or an invitation to buy or sell any securities, please see my Terms & Conditions page for a full disclaimer.