Turn messy company data into numbers a boss can act on, using code and statistics.
Click through it
Pay, the path, pros and cons and the facts on one printable page. Good for a wall or a guidance counsellor.
Or read the whole thing here
What you earn and what it costs to get there.
Median
$120,230
Bottom tenth
$67,240
Top tenth
$199,130
The national median is $120,230 a year, and the bottom tenth start near $67,240. The federal wage survey has its own Data Scientists category, so this is a clean match to the job title rather than a wider group.
About $67,000 to $85,000
Bottom tenth of the survey sits at $67,240
Very few people walk into a data scientist title straight out of college. The usual first job is data analyst, and you move across once you can code and build models on your own.
$199,130 and up
Senior and staff roles at large tech firms, banks and hedge funds
At big tech companies a lot of the money arrives as stock grants and bonuses, which the federal wage survey does not count. Real total pay at that level is often well above the published figure.
Tuition, fees, exams and kit. Not rent, food or travel
The low end is two years at a community college then a transfer to finish an in-state public bachelor's degree, with the tools costing nothing, since Python, R and the main libraries are free. The high end adds a master's degree, which many employers still expect for the title, on top of full price in-state tuition. Online master's programs change that a lot: one well known public online analytics degree totals under $12,000 for state residents. File the FAFSA (Free Application for Federal Student Aid) every year, because it is the one form that decides your Pell Grant and your federal loan offer.
Fill in the FAFSA (Free Application for Federal Student Aid) every year. It is the single form that decides your Pell Grant and your federal loan offer, and it is free to submit.
Annual bonus
Often 10 to 20 percent of base pay at a large company.
Stock
Shares that vest over four years. At a public tech company this can rival your salary.
Remote or hybrid work
One of the few well paid fields where fully remote jobs are still normal.
Conference and training budget
Most employers pay for courses because the tools keep changing.
The exact path from high school to qualified.
A degree plus proof you can actually do the work. Nobody hires from a transcript alone.
High school diploma
Grade 12
Take every math class on offer and start writing code. Python is free and runs on any laptop.
Bachelor's degree
4 years
Data science, statistics, computer science, math, economics or a science with heavy quantitative work.
Internship and projects
During college
A summer internship plus three or four finished projects on a public code profile. This is what gets you interviews.
First job with data
1 to 3 years
Data analyst, business analyst or research assistant. You learn the messy real-world part here.
Data scientist title
1 to 2 more years
Some people move up internally, some take a master's degree to make the jump faster.
Statistics
The most directly useful class you can take. Take it if your school offers it.
Calculus
Needed for the machine learning courses in a degree.
Computer science
Any language teaches you the habits. Python is the one this field uses.
English
Half the job is explaining a result to someone who does not do math.
There is no state board, no exam and no legal title protection. Anyone can call themselves a data scientist, which cuts both ways: you can get in without a specific credential, and you have to prove your skill at every interview instead. The American Statistical Association offers a voluntary Professional Statistician credential, but almost no employer asks for it. What employers do ask for is a code portfolio and a technical interview you can survive.
Dozens of universities now run a dedicated undergraduate data science major.
University of California, Berkeley
Runs one of the largest undergraduate data science programs in the country. Roughly one in five Berkeley undergraduates takes a data science class each year.
Large public universities
Most state flagship universities now have a data science or statistics major at in-state tuition.
Community college first
Two years of math and programming, then transfer into a four year university. The cheapest legitimate route.
Online master's programs
Georgia Institute of Technology and others run cheap online master's degrees while you keep working.
Things you don't need, but that help.
Machine learning engineering
Putting models into live products. Pays the most and needs real software engineering skill.
Experimentation
Designing and reading split tests for a product team. Statistics heavy, less coding.
Natural language work
Text and language models. The hottest area right now and the most crowded.
Health and biostatistics
Clinical trials and hospital data. Slower moving, very stable, often needs a graduate degree.
Fraud and risk
Banks and insurers. Excellent pay and you can explain what you do at a family dinner.
SQL (Structured Query Language)
The language for pulling data out of a database. You will be tested on it in almost every interview.
A public portfolio
Three finished projects with real data beat ten half-built ones.
A cloud certificate
Amazon Web Services and Microsoft Azure both run cheap entry level exams that get you past resume filters.
Domain knowledge
Knowing how a hospital or a bank actually works makes you far more useful than one more modeling trick.
Agencies from the Census Bureau to the Centers for Disease Control post data science jobs on USAJOBS (the federal government's official hiring site). Pay tops out lower than at a tech company, the hiring process is slow and paperwork heavy, and you may need a security clearance. In exchange you get a pension, stable hours, and problems nobody else gets to work on. Federal service can also count toward Public Service Loan Forgiveness, which cancels the rest of your federal student loans after ten years of qualifying payments.
Day to day, pros and cons, where you'd work.
Start with messages and a look at whether last night's data pipeline finished. Most of the morning is cleaning and joining data, which is the unglamorous bulk of the job. Midday is a meeting with the product or business team who asked the question, usually to find out that the question they asked is not the question they meant. Afternoon is modeling, or building a chart that answers the real question. You spend more time writing explanations than writing code.
The good
The hard parts
Things people don't tell you.
The phrase data scientist was only coined around 2008, by people building analytics teams at LinkedIn and Facebook. The federal government did not give it its own occupation code until 2018.
Survey after survey of working data scientists puts the share of time spent finding, cleaning and reshaping data somewhere around two thirds to three quarters. The modeling everyone signs up for is the small slice at the end.
Python, R, PostgreSQL and almost every machine learning library are free and open source. You can learn the entire technical side of this job on a secondhand laptop without paying for software once.
The federal occupation profile flags data science as a bright outlook career, meaning new openings are very likely and the field is expected to grow rapidly over the next few years.
Last checked September 2026. Pay figures are typical full-time annual amounts in United States dollars, based on Bureau of Labor Statistics wage data and published pay scales. They vary by state, employer and experience. Tuition is for in-state students at public schools unless the card says otherwise.