?

If you’ve been looking into a career in data analytics a common question you may have is whether you should become a data scientist or a data analyst. My experience is unusual because I was a data scientist first and then a data analyst. I realized after being in both roles that if I had been a data analyst first I would’ve had greater success in my data scientist role. Today I’d like to discuss why you should consider becoming a data analyst first and then decide if you want to become a data scientist.
Barrier to entry
The requirements to become a data analyst are lower compared to a data scientist. If you only have a bachelor’s degree you can become a data analyst by learning the necessary skills online or attending an analytics bootcamp. Becoming a data scientist will likely require you to get a graduate degree which translates to more money and time in school before you meet the requirements for a data scientist role. Data analysts primarily need to know SQL but data scientists also need to know programming, machine learning, advanced math, and statistics.
If you’re uncertain if you’ll like a job analyzing data all day then don’t spend additional money and time for an advanced degree to learn the Data Science skills. Learn SQL first and get a data analyst job to figure out where your interests lie. If you discover you want to work on machine learning problems you can continue to learn skills needed to become a data scientist.
Since I was a data scientist first I had to learn more skills in the beginning to do my job. There was a lot more pressure to get up to speed quickly. After I become a data analyst, the job was easier because I already had the necessary skills from when I was a data scientist. If I had been a data analyst first I could’ve gradually built up my skill set and reduced the stress of having to learn multiple skills at the same time.
Gaining data experience
A common problem for new graduates is how to get their first job without any experience. Since the requirements for a data analyst is lower, it’s easier to become a data analyst first to gain the analytics experience necessary to become a data scientist. If you can’t learn programming and Machine Learning on the job as a data analyst, supplement your knowledge by taking online courses and practice building models with Kaggle competitions.
As a data analyst, identify pain points that machine learning models can solve and pitch them as projects to stakeholders. These projects will provide an opportunity for you to practice machine learning and solve real business problems. For example, when I was a data analyst the sales team had two reps working full time manually going through millions of users in the company database to identify single license users more likely to upgrade to team licenses. It was time-consuming to find good prospects because reps had to look up one user at a time. I developed a machine learning model that selected users with the highest probability to upgrade for reps to contact first which helped increase conversion rates using less time. I was able to leverage a business problem to practice building a machine learning model on the job and use it as a talking point when I interviewed for my next job.
Developing soft skills
Data analysts support multiple stakeholders and present results more often because the turnaround for requests are shorter than data science projects that can require weeks to develop a machine learning model. This means as a data analyst you have more opportunities to practice developing your presentation and data storytelling skills. Learning company KPIs and how they relate to each other are key to effectively presenting actionable insights derived from your analysis results.
Since I was a data scientist first, I didn’t know to relate my model results to KPIs. This resulted in many models that were never adopted because I was unable to convince my stakeholders of their value. If I had been a data analyst first, I would’ve learned how to present my results effectively and been able to convince my stakeholders to adopt my models when I became a data scientist.
Transitioning into data science
Once you’ve worked as a data analyst and learned how to work with data, proven you can build machine learning models to solve business problems, and learned how to present your data results effectively then you’re ready to try applying to data scientist jobs.
The easiest way to transition into data science is within the same company if there’s already a data science group. You’re more likely to be accepted into a data scientist role because you’re familiar with the data, you’re a known quantity, and will require less time to get up to speed to work on actual projects. It’s time-consuming to train new employees and being able to provide value quickly is a big consideration to a hiring manager.
Another option is to apply to junior data scientist positions. Data scientist positions have high requirements because employers don’t want to hire applicants with no proven experience. However, with data analytics experience and proven business impact from building machine learning models employers are more likely to overlook the typical requirements.
Conclusion
If you can’t decide between a data scientist and a data analyst, consider becoming a data analyst first and gradually learn the skills needed to become a data scientist. This will allow you to try a data analytics role earlier to avoid spending more time to learn data science skills and more money on graduate school before realizing you made a mistake. If you end up happy as a data analyst then you’ve saved yourself time and money. Regardless of which role you decide to choose, now you know you have the option of trying both.
You might also like…
Why Data Analysts Should Apply to Data Scientist Jobs
How to Present Machine Learning Results to Non-Technical People