Careers in Data Science: Data Analyst vs. Data Scientist
The most in-demand resource of the 21st century is data. Analyzing data allows businesses, governments, institutions, and individuals to make more informed decisions in our increasingly complex, interconnected world. More data is being collected now than at any point in human history. Most of this data hasn’t yet been analyzed.
Demand for data experts continues to increase at a staggering rate. The number of data analytics jobs has grown by 650% since 2012. The U.S. Bureau of Labor Statistics estimates that over 11.5 million new data jobs will be created by 2026.
As the data landscape has grown, many new jobs and positions have sprung into existence. Two job titles you’ve likely heard about are data scientist and data analyst. Although they sound similar, there is a difference between them. What skills, education, and experience do you need to work in these roles, and how much can you expect to make in each one? Read on to find out!
Role responsibilities of a data analyst
The primary responsibility of a data analyst is to derive insights from data. The process they use has three stages: extraction, manipulation, and presentation. An analyst queries data, contorts it into the proper shape for analysis, and presents their conclusions. A skilled data analyst possesses both technical skills and business acumen. They need to know not just how to crunch numbers but which numbers to crunch and what they mean for the business. Their goal is to explain the why and the how simply and effectively.
In the extraction stage, the analyst acquires a data set from one or more sources. This may involve ingesting a file of a certain format like CSV, XLSX, or JSON. It may involve querying a database for the correct data by specifying the right combination of filters. It may involve joining multiple data sets from different sources together.
In the manipulation stage, the analyst contorts or shapes the data into a format suitable for a calculation. The calculation does not have to involve a complex statistical operation. For example, a business stakeholder may only need a count of records or an average of numbers to make an informed decision. A data analyst seeks to answer the question, “How do I get the data into the right shape so it can reveal its hidden meaning?”
In the presentation stage, the data analyst determines the best way to “tell a story” with the data. Sometimes that is as simple as delivering a single conclusive number to a stakeholder. Other times it may involve creating a spreadsheet with a subset of important rows from a larger data set. In some cases, you can include visual aids like bar graphs, pie charts, or histograms. Regardless of the medium, the analyst seeks to present the conclusion compellingly.
A data analyst working at Udemy may answer questions like:
- What were the five top-selling courses in each category last month?
- What was the average revenue for all photography courses in the last year?
- What percentage of a course does the average student complete?
A data analyst can work with various technologies, including:
- Graphical spreadsheet applications like Microsoft Excel or Google Sheets
- Generalist programming languages like Python, Java, or Scala
- Statistical programming languages like R
- Computing environments like MATLAB or SAS
- Data analysis libraries like Pandas
- Relational database management systems like Postgres or MySQL
- Data visualization software like Tableau or Power BI
As a data analyst gains experience, they learn which tool is best for each job. There is rarely one “perfect” solution. Rather, each tool has its own advantages and disadvantages.
Role responsibilities of a data scientist
The key distinction between data analysts and data scientists is that the latter build predictive models. A data analyst aims to understand what has happened. A data scientist seeks to predict what will happen.
There are several areas of specialization within the data science ecosystem. For example, machine learning focuses on teaching computers to recognize patterns based on experience. On a site like Udemy, machine learning models help predict what courses will be appealing to you based on what courses you’ve already purchased.
Another example is natural language processing (NLP), which focuses on programming computers to process and understand written text and human language. On Udemy’s site, NLP models can generate automatic subtitles for videos by identifying common patterns in human communication.
There are additional specializations in data science, including artificial intelligence, deep learning, cloud computing, and more.
Data scientists build predictive models based on existing data. This additional requirement requires additional technical skills. A data scientist’s background may include experience in:
- Software development, programming, and computer science
- Math, statistics, and complex algebra
- A specialized domain like physics or finance
A data scientist is likely to have experience in a greater breadth of technologies than a data analyst. They may work with:
- Distributed processing solutions like Spark or Hadoop
- Machine-learning software like TensorFlow or PyTorch
- Cloud storage solutions like AWS, GCP, or Azure
Data scientists may find data engineering duties in their job requirements. Data engineers build the pipelines that collect and store data. They have to consider questions such as:
- Where is the data going to be stored?
- How are we going to connect to it?
- What size is the data?
- What is a suitable software solution for that scale?
- How is the data being stored?
- What details can be discarded?
- What details must be modified?
- What details can be kept in their original format?
Certain professions define concrete boundaries for roles. For example, an otolaryngologist is a doctor who diagnoses ear, nose, and throat problems, while a podiatrist is a doctor who treats conditions with the foot and ankle. Each doctor has their domain of expertise.
The world of data is a lot more fluid. If you search for “data” jobs, you’ll find roles for a data analyst and a data scientist alongside other titles like data engineer, data architect, big data developer, statistician, and business intelligence analyst.
The responsibilities of these roles often overlap and can vary from market to market, from company to company, and from team to team. When interviewing for a data position, a candidate should ask what types of projects their team will be tackling. A “data scientist” can mean something completely different across two different companies.
Salaries and education
The job search engine Indeed estimates that data scientists have the highest average education level of any job in today’s marketplace. 75% of data scientists have advanced degrees (master’s or PhDs). A data analyst is half as likely to have such a degree. The barrier to entry is thus significantly higher for data scientists. With that said, changing professions from data analyst to data scientist is a common career path as one acquires higher education.
Due to the shortage of workers, high demand, increased complexity of their roles, and employer expectations of higher education, data scientists typically earn more than data analysts.
Job review site Glassdoor estimates the average annual salary of a data analyst in the United States to be $62,453. Salaries vary from market to market. For example, the average data analyst salary in San Francisco, California, is $84,658 but only $52,939 in Oklahoma City, Oklahoma.
In comparison, the average salary for a data scientist in the United States is $113,309. The median entry-level salary for a data scientist is $95,000, which is the highest entry-level salary of any role in the United States. If we do a similar comparison between cities, a San Francisco data scientist can expect to make around $140,897 while an Oklahoma City data scientist can expect $90,512.
A clear salary gap emerges between the two roles, even when accounting for market differences. The gap is reflective of the higher credentials and technical skills that data scientists typically have.
All of us have done some data analysis or data science in our lives without even knowing it. Maybe you’ve crunched your monthly expense numbers to calculate your spending budget. Or perhaps you’ve predicted how your favorite athlete will perform based on their past performance.
In these situations, we’re using data to observe what has happened and make predictions on what will happen. Data analysts and data scientists do the same thing in their day jobs but with a lot more technological and statistical savviness. There are many technology courses available on Udemy if you’re interested in learning more about the tools they work with. Best of luck in your data journey!
Top courses in Data Analysis
Data Analysis students also learn
Empower your team. Lead the industry.
Get a subscription to a library of online courses and digital learning tools for your organization with Udemy Business.