What is the difference between a Data Analyst and a Data Scientist?
There’s a bit of a confusion as what the difference is between a data analyst and a data scientist.
I went over 100 job descriptions and I must have read over 50 different articles and white papers from reputable sources, including top-notch universities. I thought I knew what the difference between these two were, but let me tell you what I gained from researching on what others thought the differences were.
I would like to preface that there’s not one role that is better than the other. What I would like you to take away from this article is how one role or the other would make more sense for the skills that you have or want to gain and what you would be more happy working as. In the end, that’s what’s most important, isn’t it?
Let’s go into it and cover the, SCOPE, SKILLS, EDUCATION & EXPERIENCE, and SALARY EXPECTATIONS for a data analyst and a data scientist and we’ll see which one is the right for you.
What I learned from 100 job descriptions
As I mentioned, I read over 100 job descriptions from LinkedIn.com, Indeed.com, Glassdoor.com and whatever I could find on google. What did I gain out of it? Total confusion!
Some were saying that the data analyst deals with big data frameworks and systems (nah, I don’t think so), or data scientists build data pipelines (isn’t that the role of a data engineer?), or that data analysts just cleanse data and build reports (what?). Not that’s anything wrong with that, we need more of that, but no, that’s not exactly what a data analyst does.
Some articles and whitepapers made things even worse. There was a total mish mash between the responsibilities of a data scientist, a data analyst, a data engineer, and even a business intelligence programmer.
There were also quite a few data sources that had overlaps as to what the responsibilities and skills of a data analyst vs a data scientists were. So let’s start making some sense out of that so you can see what role would be a better fit for you.
The scope of work for a data analyst vs. a data scientist
The scope for a data analyst is said to be micro as they usually work with static data, so a snapshot of that data, and it generally tends to be structured data. It’s limited to mostly using statistical tools and techniques, usually in the realm of descriptive analysis.
The data analyst analyzes data to answers specific questions, such as: “why sales dropped in a certain quarter”, “why a marketing campaign fared better in certain regions”, “how internal attrition affects revenue”, “where budget spending should be increased to get more return on investment”, “what’s the number of leads that the sales efforts must generate to fill the sales pipeline”, and so on.
This role analyses structured data, usually through the use of descriptive analytics, to answer questions generated for better business decision making
On the other hand, the scope of data scientist is said be macro as they work more with dynamic data that can change frequently and whatever models and solutions they’ve built needs to adapt on the fly. So yes, they are more likely to work with big data that’s both structured and unstructured and that can have a high volume, velocity and variety. The data scientist’s scope might also involve artificial intelligence, machine learning and deep learning.
The data scientist, through data analytics, defines new business questions and problems that lead to innovation and then uncovering the solutions to these new and unknown problems. Contrary to the data analyst, the data scientist tends to already understand the company’s requirements in order for them to formulate new questions that need to be addressed. These questions can sometimes be addressed by data analysts, by the way.
The data scientist works with structured and unstructured data and uses various machine learning tools and algorithms to forecast and classify patterns in data and increases the performance and accuracy of these algorithms. So more on the predictive analytics and modeling.
Some scope overlap of both roles:
Both roles tend to do some data cleansing and data transformation - and it’s probably part of the job that both roles complain about. The difference being that a data scientist tends to work with more complex data and raw data. By the way, from the data science point of view this could fall under data wrangling or data munging.
They both also present and communicate their findings through the use of data visualization and data storytelling. There are sources that state this is more in the scope of one role over the other, but I think it belongs to both.
The skills needed for a data analyst vs. a data scientist
The data analyst needs to know statistics and have a mathematical aptitude, but I would rate the level of this skill as intermediate. Since the data analyst works with structured data, they need to understand relational databases and have knowledge of SQL and should be well versed in using Excel (believe it or not).
The data scientist needs to have advanced knowledge of statistics and math. From what I find, most data scientists tend to gravitate towards mathematics, though. Of course, there are superstars that excel at both. The data scientist should also know SQL, but also NoSQL such as MongoDB and Cassandra as these NoSQL databases have the ability to handle and scale dynamic data, big data.
As for programming languages, data analyst should have a basic knowledge and data scientist have advanced knowledge of Python, R, Julia (which is an upcoming language that’s being touted as the successor of Python), sometimes SAS (which is commonly used in the banking and financial sectors), maybe even Jupyter Notebooks — which is not a language, but a web application that you can use to create and share documents that contain live code, equations, visualizations for your data analytics and data science projects. So all of these really depend on the company and industry that they would work in and the technical environment they have in place.
A couple more things to add to the data scientist, if their focus is on deep learning they might need to know of specific deep learning frameworks (ex: TensorFlow, PyTorch, Keras, Caffe)or if their focus is on artificial intelligence, they might need to know some auto machine learning tools (ex: AutoKeras, IBM Watson, DataRobot, Amazon’s Lex).
Both of these roles should know how to utilize data visualization tools (ex: Tableau, PowerBI, and Qlik being the most popular). Lastly they should both be good at problem solving and possess communication skills, including data storytelling.
The education and experience needed for a data analyst vs. a data scientist
As I mentioned, I looked at over 100 job descriptions and the requirements can vary greatly. Here are some common denominators, though.
The requirements of a data analyst are for a Bachelor degree at a minimum, desirably in a science field. And it’s not uncommon to see that it’s desirable to have a statistics or mathematics background.
On the other hand, for a data scientist in most cases it is required having a Masters degree, usually in mathematics, or software engineering, machine learning, or computer programming. Some even require for a Ph.D.
As for the years of experience as a data analyst you could pass by with 2 years of previous experience and as a data scientist with 4 or more. These can oscillate depending on who is writing the job description, but I found these to be the most common minimum requirements.
The salary expectations for a data analyst vs. a data scientist
And as there’s a difference in the skills and experience required as well as the scope of the work, the same can be also be said for the salary for both these roles.
As of April 2021, the average salary of a data analyst in the US according to Glassdoor is around $66k-$67k a year. It should come as no surprise, given the scope of work of the data scientist role that they earn significantly more money than their data analyst counterparts. Again, Glassdoor has the average at $113k a year.
Which role right for you? Data analyst or data scientist?
It ultimately comes down to what you want to do and what skills and experience you want to pursue. Whilst one role does earn a higher salary and requires a more in-depth skill set — neither can exist without the other and both play integral roles for organizations when it comes to their data. So which role do you think is right for you?