Transitioning From Data Analyst to Data Scientist

Data nerds of the world, rejoice! Data science has topped “best jobs” lists for years, and those in data science report high salaries and job satisfaction. All signs point to unmet (and growing) demand. But you knew that didn’t you?

As a data analyst, you’re ahead of the game. You’re tech savvy and you’ve got top-notch critical thinking skills. But making the transition from analyst to data scientist isn’t a slam-dunk. Even the best data analysts will have to break out of their comfort zone in order to prove themselves as data scientists.

For a breakdown on the differences between the two, check out our guide here. If you’re already clear on the differences, read on: here are five critical steps to follow to transition smoothly.

Network Smartly

Data scientists are always looking for the biggest lever. They’re thinking about long-term growth - and they know that no department operates in a vacuum. Every function is connected.

Your perspective is the biggest lever you have to build a data science mindset. Learn to think big-picture. This means learning about different business functions. Make friends in R&D, accounting, and HR, and find out what drives them crazy. What intranet sites crash? Do they read auto-generated reports? Which KPIs are they tracking daily? Answering questions like these is the fastest way to get a sense of what’s working and what’s not, and if you can come up with solutions you’ll be noticed.

This approach comes with an added bonus. You’ll get lots of practice speaking to non-techies. As a data scientist you’ll work with high-level management who will want you to speak plain English. If you can explain your ideas to your accounting friend, in time you’ll be able to do the same with the VP of accounting.

Take on Stretch Assignments

Once you’ve made friends in other departments, it’s time to get to work. You’ve identified some pain points, and with your data analyst’s toolkit you’re uniquely positioned to help. Create a dashboard for the accounting team tracking most frequently used metrics. Help the HR team identify trends in exit interview data.

There are a million ways to customize the data you have to better serve your internal customers. How will you choose which project to take on? Ask yourself what you’d want to know if you were VP of X department. Soon, inter-departmental patterns will start to emerge and you’ll see higher-level solutions. At this point you’ll know you’ve leveled up in your quest to become a data scientist.

Position Yourself as an Authority on the Latest Tech

You’d be shocked at the complexity involved in a software investment decision at a big company. Sales cycles can take years of committee discussions and cost millions. No one will expect you to swoop in with a perfect solution, but you can create huge personal value by positioning yourself as a go-to for small tech decisions and fixes.

Your organization probably doesn’t even know what’s available out there. Study up on the newest innovations, sign up for a few trial offers, and offer suggestions based on your newfound experience. Bonus points for dabbling in truly cutting-edge tech like AI - people will associate you with the latest and greatest, strengthening your position when you apply for that data science role.

Add the Right Tools to Your Toolbox

Communication skills, storytelling ability, and problem-solving are important facets of data science but you’ll also need to expand your technical skills. You’ll need to know a few programming languages (Python, R, MySQL, etc.) and you’ll need a thorough education in statistics and math. As a data scientist you won’t just be analyzing trends - you’ll be running experiments, building models, and chasing down answers to business problems.

This part of your transition plan should be super fun! Let your intellectual curiosity lead you. Take on passion projects that necessitate some stats or that new language. Interested in conservation? Data.gov has some great agricultural data sets. Motivated by time pressure and have a competitive streak? Enter a Kaggle competition. Try your hand at machine learning. Create some sleek visuals to go along with your findings. And document everything.

Some of you are probably wondering about formal education. It’s true that many data scientists have advanced degrees. By Bureau of Labor estimates, about a third. But diving into a program mid-career when you already have real world experience with data might push you off-track. You’ll spend anywhere from 2 to 7 years getting that degree, and you won’t exactly have tons of free time to keep up on all the cool tools and trends developing. Only you can make this decision—but weigh your options carefully.

Which brings us to the final step.

Showcase Your Hard Work

You’re a go-getter with a million ideas to test and the skill set to produce results. Your portfolio should reflect that.

Choose a few projects that match up with the specific job you’re after. Mix it up - share a data cleaning project, a visualization project, and an exploratory analysis. It’s best to choose the work you’re most interested in - you’ll be more comfortable going in depth during an interview, and your passion will shine through. Other great samples include:

• Those Kaggle competitions! Bonus points for any team experience - you want to showcase yourself as a team player and a clear communicator.
• Hackathons/Sprints. Look for developer conferences or check with your local tech meetup for these. Sprints in particular are great for an aspiring data scientist because you’ll learn the fundamentals of Agile methodology, which you’re bound to run into if you’re not already actively involved.
• Open-source contributions. Passion-projects are great, but working on open-source can teach you how to write production-level code. Welcome to the big time.

Crush the Interview

If you're a data analyst now and you’ve followed the above steps, you’re probably a strong candidate already. Congratulations! But you’re not the only one attracted to the perks of data science. You won’t be the only strong candidate out there. A few final touches can help you seal the deal.

• Hone your storytelling ability. If you’re applying at a new company you’ll need to spend a little more time on translating your inter-departmental projects for interviewers who won’t have any of the context. Make sure to cover the problem, your solution, and the results without too much fluff. Your hiring manager will probably be comfortable with tech lingo, but they’ll be looking at your communication skills too. A clear, coherent, and interesting story will leave a lasting impression.
• Practice, practice, practice. There’s no substitute for a technical interview besides practice. Work as many problems as you can find, in conditions similar to the interview. If you can, get your friends to act as interviewers and don’t be afraid of constructive criticism. When the time comes, you’ll be ready.
• Don’t forget to prepare your questions for your interviewer. It’s easy to get swept up in technical preparation; we’ve all been there. But don’t forget to think about what you’d like to ask your hiring manager. Data science is still fairly vague, so make sure you walk away with a sense of the company culture and what you can realistically expect on a day-to-day basis. You’ve done so much work to prepare it only makes sense to work for a company that’s a good fit.