Role profile
Data scientist
Also called: Machine learning engineer (overlaps) · Applied scientist · Data analyst (senior, overlaps) · Quantitative analyst (overlaps) · Decision scientist
Data scientists use statistics, programming and machine learning to answer business questions and build predictive models from large datasets. They are hired by tech companies, consulting firms, insurers, banks, retailers, healthcare organizations and government agencies.
- Median pay, 2025
- $120,230
- $67,240 to $199,130 (10th to 90th percentile)
- US jobs, 2025
- 275,600
- Data Scientists
- Outlook, 2025-35
- +35%
- Much faster than average
- Typical entry
- Bachelor's degree
- No prior experience required
What the work looks like
- Pin down what question a stakeholder is actually asking before touching any data.
- Write SQL to pull and join data, then spend a large share of the time cleaning and checking it.
- Build and evaluate statistical or machine learning models, and compare them against a simple baseline.
- Design and analyze A/B tests, including telling people when a result is not significant.
- Turn findings into charts, notebooks and short write-ups that a non-technical manager can act on.
- Work with data and software engineers to put a model into production and monitor it afterward.
- Evaluate when a large language model or other off-the-shelf AI tool is good enough and when a custom model is worth the effort.
How people get in
The BLS lists a bachelor's degree in math, statistics, computer science or a related field as typical, and notes that some employers require or prefer a master's or doctorate. In practice, many data scientist postings ask for a graduate degree, while analyst roles are more open to bachelor's holders.
A common path is to start as a data analyst, get strong at SQL and business questions, then add statistics and machine learning on the job. People with quantitative degrees in physics, economics or engineering also move in, usually through a master's program or a portfolio of real analyses.
Some industries want domain knowledge: finance employers may expect coursework or experience in investments, healthcare employers familiarity with clinical data. A portfolio that shows a full analysis, from messy data to a clear recommendation, is worth more than a list of courses.
Skills employers ask for
- Python (pandas, scikit-learn)
- SQL
- Statistics and experiment design
- Machine learning fundamentals
- Data visualization
- Jupyter notebooks
- Git
- Cloud data warehouses (BigQuery, Snowflake or Redshift)
- Communicating results to non-technical audiences
- Basic data engineering (pipelines, scheduling)
Outlook to 2035
The BLS projects 35% growth from 2025 to 2035, the fastest of the roles we track, with about 24,800 openings a year, as organizations want more data-driven decisions. The title covers a wide range of work, from dashboards to production machine learning, so postings vary a lot in what they expect. The projection is strong, but entry is competitive and graduate degrees are common among those hired.
How the pay compares
Data scientist highlighted against the other 13 roles we track.
Source: U.S. Bureau of Labor Statistics, Occupational Outlook Handbook, Data Scientists. Pay is 2025 median annual wage; outlook is projected employment change 2025-35.