For three years running, career site Glassdoor has named the same job the ‘Best Job in America’: data scientist. With an average base salary of $110,000, high job satisfaction, and a very strong demand in the job market, it’s not hard to see why it consistently tops the rankings.
To me, data science is the new kale. A few years ago I’d never even heard of it, and now, it’s everywhere. But how many of us actually know what a data scientist does?
I decided it was time to find out, so I spoke with Kyle McKiou, founder and chief data scientist at Data Science Dream Job. A former lead data scientist at Anheuser-Busch InBev, McKiou has held several other data science roles as well.
Our conversation touched on what the job entails, the pros and cons of the role, and McKiou’s tips for someone hoping to secure one of these coveted positions.
The Day-to-Day: What Does a Data Scientist Do, Anyway?
DH: What does a data scientist do on a day to day basis? What are their main responsibilities and goals?
KM: The main responsibility of a data scientist is to solve business problems. More specifically, the data scientist uses data, mathematical techniques (e.g., statistical and machine learning models), and computing to solve problems. Typically, the solution involves automating or optimizing an existing process or completing an analysis that creates actionable insights and will increase revenue or decrease costs.
It’s a very applied, “hands on,” role and it doesn’t focus primarily on theoretical research or algorithm design, despite what some may think.
What we do on a day-to-day basis varies a lot, but it generally includes:
- Meetings with other project stakeholders, IT, product teams, and engineering teams.
- Internal team meetings to discuss our work, progress, and approach.
- Meetings with clients (internal or external) to get project feedback.
- Data cleansing and exploration.
- Automating processes.
- Integrating processes.
- Researching techniques and technologies that may be useful.
- Building and testing machine-learning models and methodologies.
- Integrating business rules and requirements into complex mathematical models.
- Writing and testing code and data pipelines.
The Pros and Cons of Working in Data Science
DH: What do you like most about the job?
KM: It’s a very unstructured job, which leads to a lot of possibilities. The exact approach, tools, and technologies you use to solve problems is up to you. There is nothing that says you have to do things a certain way. This is awesome because it always keeps the job challenging and interesting.
DH: What’s your least favorite aspect?
KM: It’s not a well understood role, which means that you’ll have to spend a lot of time educating other people within the company (and clients) on what you do and why it’s different than the analytics that they’ve been doing for years.
On the Sudden Popularity of Data Science
DH: Why do you think Data Scientist has become such a coveted job?
KM: It’s a challenging, exciting job that doesn’t have a lot of boundaries. Also, it can make a massive impact on the bottom line for a company.
A lot of people that come from highly technical backgrounds, especially those stuck in academia, are sick of doing incredible work and then seeing very few, if any, people or companies leverage their results to make an impact. In data science, you get to see your results put into the marketplace to drive real change, which is extremely rewarding.
For example, if you build a model that allows a company to better target customers and this decreases acquisition costs by just 1%, it could mean millions of dollars saved for the company. Also, it will result in happier customers that now see more relevant messaging instead of being bombarded by irrelevant marketing that they don’t care about all day.
It’s win-win – you get to make life better for your employer and the customers while doing something that is both challenging and interesting.
Tips for Breaking into the Business
DH: What skills should you develop if you want a data science job, and what programming languages should you learn?
KM: First off, you need to focus on the foundational math, stats, and computer science before diving into all the machine learning algorithms.
As far as programming, Python is the best language to learn, but R is appropriate for some roles and companies. Also, SQL is used pretty much across the board but isn’t always a “make it or break it” skill.
DH: How did you decide to get into data science?
KM: I read an article about how it was a “sexy” job where a massive impact could be made and I knew that it was what I wanted to do. Unfortunately, I really struggled to break into the field just like everyone else.
Through months of trial and error, I was able to figure out what companies were looking for when it came to hiring data scientists. I realized it was all about being able to demonstrate that you’re capable of doing the job of a data scientist and presenting yourself and your work in a way that is compelling and relatable to companies.
DH: Do you have any advice for would be data scientists who are trying to break into the industry?
KM: Don’t get overwhelmed by crazy job descriptions or the amount of material that it seems like you need to know. Everybody else is facing these exact same challenges. If you’re interested in a job, then apply for it right now. Push your job search forward by taking action immediately and repeatedly until you are able to identify your weaknesses and then attack those while you continue to apply.
DH: What do you think about tech “boot camps” for data science that purport to teach you everything you want to know in a few months?
KM: I don’t typically recommend boot camps since most of them don’t prepare people to actually get jobs. Plus, there are less expensive ways to learn the technical skills.
Personally, I’m happy I got a chance to talk to Kyle. Now, when I meet a data scientist, I no longer have to secretly think, “I have no earthly idea what you do.” It turns out that data scientists are like most other analysts — they’re just highly trained to use specific tools.
If you have an analytical mind, a technical background, and a strong desire to work in a cutting-edge field, it seems like a path worth exploring. If you’re interested in data science as a career, you can follow McKiou on LinkedIn or sign up for his newsletter. He uses both platforms to offer tips to aspiring data scientists.