Data Analyst vs. Data Scientist: What’s the Difference?

If you’ve ever gone through a job search in the data/analytics world, you’ve probably seen these two job titles (and slight variations on them) quite a bit: Data Analyst and Data Scientist. But what exactly is the difference between these two roles?

The short answer is it depends. In reality, these job titles are quite fluid. A Data Scientist at one organization might have roughly the same job description as a Data Analyst at another organization. Sometimes a Data Analyst may “cross-over” and do work typically reserved for a Data Scientist, or vice versa.

On top of that, there are other job titles that, depending on the organization, can overlap partially or completely with typical Data Analyst or Data Scientist roles. For example, a Product Analyst at a tech company could effectively be a Data Analyst that primarily focuses on product-related work.

That said, we can still draw some general distinctions between Data Analysts and Data Scientists. None of these distinctions will be true everywhere, but hopefully they’re enough to help you determine whether you’d rather pursue a Data Analyst or Data Scientist career path, or whether you should hire a Data Analyst or Data Scientist, if your organization is looking to incorporate data into your product or decision-making process.

What do Data Analysts and Data Scientists typically do?

Data Analyst: Use data to generate insights that can help the organization make better decisions. For example:

  • Analyze a product’s conversion funnel
  • Create a dashboard to track business performance
  • Forecast next year’s revenue
  • Compile and analyze market share trends
  • Analyze a quarterback’s passing tendencies

(I’m using examples from tech and sports because, well, that tends to be most of my audience. That said, the core skillsets of Data Analysts and Data Scientists have applications across a number of other fields)

Data Scientist: Build, maintain, and improve models and algorithms that, in many cases, become a part of the organization’s product. For example:

  • Build a machine learning model that optimizes a site’s homepage based on a user’s known characteristics
  • Backtest two different models that set betting lines to determine which model performs better
  • Develop a predictive model to estimate the long-term value of NFL draft prospects

What technical skills do you need for these roles?

Data Analyst: Between the two roles, Data Analysts tend to work on less technical projects. Because of that, you’re more likely to see folks without highly technical degrees in these positions (e.g. There’s probably a higher incidence of Data Analysts with Economics or Business degrees than Data Scientists with Economics or Business degrees).

That said, working with data naturally requires some technical skills. Here are a few of the most important for a Data Analyst:

  • Most organizations with large databases use SQL to retrieve the data.
  • Excel (or Google Sheets) gets some flack because it’s not great for more advanced statistical analysis and modeling, but it’s still an incredibly powerful tool for doing quick analyses and creating visualizations.
  • R is a programming language built specifically for statistical analysis, and is a bit easier to learn than Python.
  • Many organizations use data visualization software like Tableau or Power BI to create reporting dashboards that can easily be accessed across the organization. Data Analysts are often building those dashboards.

Data Scientist: Data Science can be a highly technical profession, with a skillset that begins to resemble a Software Engineer’s. Because Data Scientists are often building models that will become a part of the product, they typically need to be more comfortable interacting with the product’s codebase. Here are some of the most important technical skills to do that work:

  • SQL, for the same reason stated above.
  • Python is great for building advanced models, and can typically integrate with a product’s codebase in a way that R can’t. The learning curve for Python is steeper than for R, but Python can do quite a bit more.
  • That said, it can’t hurt to know R as well ¯\_(ツ)_/¯

What non-data-centric jobs do these two roles resemble?

Data Analyst: Business/Strategy Consultant

Data Scientist: Software Engineer

How much do folks in these roles get paid?

Data Analyst: Probably less than a Data Scientist of equivalent experience.

Data Scientist: Probably more than a Data Analyst of equivalent experience.

High tech skillsets tend to demand higher salaries, so the more technical nature of data science means Data Scientists will tend to get paid more, all else equal.

Anything else to know?

Because there’s a lot of overlap between these two roles, it’s not unusual to move between them. I’ve known Data Scientists who later became Data Analysts and vice-versa.

Similarly, the skillsets for both of these roles are transferable across industries. That means you can cut your teeth in sports analytics and transfer those skills into data science at a tech company. Or you can get your start as a Data Analyst in academia and later move into a quantitative role on a political campaign.

Either way, you’ll have highly transferrable skills and experience, and that’s tough to beat.