We earn a commission if you buy through links on this page. How we're funded.
Quick answer
Data scientist salary is determined mainly by industry, company size, seniority and how close your work sits to revenue decisions. The U.S. Bureau of Labor Statistics tracks data scientists as a distinct occupation: their median annual wage was $120,230 in May 2025, and it projects employment to grow 35% from 2025 to 2035, against 3% across all occupations. The spread between entry-level and senior pay is substantial, and this guide explains what drives it. If you are starting the path from scratch, DataCamp's Associate Data Scientist in Python is the course we would start with: about 90 hours for beginners, from your first line of Python to scikit-learn models.
See Associate Data Scientist in Python on DataCamp →
What is a data scientist actually paid for?
A data scientist is paid to turn messy data into decisions the business will act on. That includes framing the question, building the dataset, modelling, and communicating a result clearly enough that someone changes what they do. The modelling is often the smallest part.
Compensation tracks influence. A data scientist who produces well-crafted analyses nobody uses is paid less than one whose work changes pricing, targeting or product decisions, even if the technical work is comparable. Understanding this early is the single most useful thing for your earnings trajectory.
The U.S. Bureau of Labor Statistics Occupational Outlook Handbook publishes median wages, typical entry requirements and employment projections for data scientists, and is the most reliable public reference point for the occupation.
Which factors drive data scientist salary the most?
Data scientist salary is driven most by employer type and by whether the role is engineering-weighted or reporting-weighted. Location, seniority and domain expertise follow. Certifications and degrees mostly influence which interviews you reach.
The table below compares Impact on pay and Explanation across 7 factors.
| Factor | Impact on pay | Explanation |
|---|---|---|
| Employer type | Very large | Technology firms, finance and large consumer platforms pay well above the occupational median |
| Role weighting | Large | Engineering-heavy data science pays more than dashboard and reporting-heavy roles |
| Business proximity | Large | Working on revenue-critical questions raises both visibility and compensation |
| Location and remote policy | Large | Location-adjusted bands versus single national bands can differ considerably |
| Domain expertise | Moderate | Regulated fields such as healthcare, insurance and risk reward subject knowledge |
| Advanced degree | Moderate early, small later | Common entry route, but overtaken by track record within a few years |
| Certifications | Small and indirect | Useful for screening, especially for career changers, rarely change the band |
How does data scientist pay differ by seniority?
Data scientist pay rises sharply from entry to mid level, then depends on whether you move towards technical depth, product influence or management. The role has an unusually wide band because the title covers everything from routine reporting to applied research.
Entry level
Entry-level data scientists are frequently hired into roles that are analytics in practice. Pay is closer to analyst rates, and the fastest way up is to take on the modelling and engineering work nobody else wants.
Mid level
Mid-level pay reflects independence: can you scope a problem, choose an approach, and defend the result to a sceptical stakeholder without supervision? This is where the analytics-weighted and engineering-weighted paths visibly separate in compensation.
Senior and principal
Senior data scientists are paid for judgement and for preventing expensive mistakes, such as building a model where a simple rule would do, or shipping an analysis with a subtle selection bias. Principal-level pay generally requires organizational influence beyond your immediate team.
The Stack Overflow Developer Survey reports compensation broken down by role and years of experience, which makes it a practical sanity check against inflated online figures.
Why do published data scientist salary figures disagree?
Published data scientist salary figures disagree because the title is applied inconsistently across the market. One company's data scientist writes production machine learning code; another's builds weekly dashboards. Averaging both produces a number that describes neither.
Other reasons the figures diverge:
- Base salary versus total compensation. Equity-heavy packages at large technology firms are not comparable to cash-only packages elsewhere.
- Self-selection. People who report salaries online skew towards high earners in expensive cities.
- Currency and cost-of-living differences that are often stripped out of viral comparisons.
- Stale data. Aggregators may blend submissions collected over several years.
The practical response is to treat any single figure as a hypothesis, then test it against official data and people doing the actual job in your market.
Not sure this is the right one for you?
Tell the picker about your background and what you want the certificate to do, and it narrows the list to the one or two courses we would start with. It suggests only our affiliate partners’ courses, and says so before it suggests anything.
Try the AI Certification Picker →How does data scientist salary compare with related roles?
Data scientist salary usually sits above data analyst pay and slightly below machine learning engineer pay, with heavy overlap in all directions. The differences reflect how much engineering the role demands rather than how difficult the thinking is.
- Versus data analyst: analysts typically earn less, though senior analytics roles at data-mature companies can exceed junior data science pay. Our data analyst vs data scientist comparison sets out the differences in detail.
- Versus machine learning engineer: engineering-weighted roles usually pay somewhat more, as covered in our machine learning engineer salary guide.
- Versus analytics engineer: a growing middle path, generally paid between analyst and data scientist levels.
Our view: the more reliable way to move up a band is to add the missing engineering skill to an analytics background, rather than to add more statistics to an already statistical profile.
Do certifications increase data scientist salary?
Certifications rarely increase data scientist salary directly, but they can be decisive in getting a first interview when you have no relevant job history. Their effect fades quickly once you have delivered work an employer can point to.
Where they help most:
- Career changers who need a structured curriculum and a credential that a recruiter recognizes.
- Enterprise and public-sector employers with formal screening criteria.
- Cloud-aligned data roles, where credentials in the Microsoft credentials catalogue or on the AWS certification site match the platform in use.
- Applied AI skills, where programs listed by IBM Training or a foundational specialization give a defensible baseline.
Our certifications for data scientists guide compares the main options, and the certifications for data analysts guide is the better starting point if you are earlier in the pipeline. In both cases, a free course paired with a genuine project usually outperforms a paid certificate with no portfolio behind it.
For a career changer who wants one structured curriculum, DataCamp's Associate Data Scientist in Python is the one we would start with: 23 courses and about 90 hours, from your first line of Python through pandas, a five-course statistics block and three scikit-learn courses. Know its limits before paying. Machine learning is only its last three courses, it teaches no SQL, and finishing it earns a completion record, not DataCamp's separately assessed Data Scientist Associate certification, so it moves your salary only through the projects you build with it.
Which industries pay data scientists best?
Technology, quantitative finance, and large consumer platforms generally pay data scientists best, because data work maps directly onto revenue there. Healthcare, insurance and energy pay competitively for domain-experienced candidates. Non-profits, education and much of the public sector pay less but often offer stability and clearer boundaries.
Sector choice also shapes the work itself, which affects long-term earning power as much as the starting salary:
- Technology and consumer platforms offer scale, experimentation culture and mature tooling, which builds a transferable track record quickly.
- Finance and insurance offer rigorous modelling and regulatory scrutiny, which develops depth that is difficult to acquire elsewhere.
- Healthcare and life sciences offer unusual datasets and long project horizons, with domain knowledge that compounds in value over time.
- Public sector and non-profit roles offer stability and mission alignment, at the cost of slower tooling and narrower salary bands.
Skill demand shifts by sector, and the Coursera Job Skills Report is a useful signal for which capabilities employers are actively adding to job requirements. Reading current postings in your target industry remains the most accurate approach.
How should you research your own data scientist salary range?
Research your own range by anchoring to official data, adjusting for role weighting, and validating with people in comparable jobs. Do this before, not during, a negotiation.
- Establish the occupational baseline from the Bureau of Labor Statistics Occupational Outlook Handbook.
- Adjust for whether your target role is analytics-weighted or engineering-weighted.
- Cross-check against the Stack Overflow Developer Survey for your country and experience level.
- Collect published bands from job adverts in jurisdictions that require pay transparency.
- Ask the recruiter for the band before technical interviews begin.
Write down the range you conclude is realistic, along with the evidence behind it, before you enter any conversation about money. Candidates who have done this negotiate calmly from a documented position; candidates who have not tend to anchor on whichever number they saw most recently online.
Who should not pursue data science for the money?
You should not pursue data science mainly for the money if you dislike ambiguity, stakeholder management or spending most of your time on data cleaning. The advertised salary is attached to a job that is far less glamorous than its reputation.
Consider a different path if:
- You want to write software full-time. Software engineering pays comparably with more predictable work.
- You dislike persuading people. A large share of the job is convincing others to act on a finding.
- You want fast feedback. Many projects take months and some are abandoned before delivering value.
If the field still appeals, our how to become a data scientist guide sets out a realistic entry path.
Ready to start?
Included in a DataCamp subscription rather than bought outright. DataCamp's pricing page shows the plans and the price for your country, and one subscription covers the rest of its catalogue too.
Frequently asked questions
Is data science still a well-paid career?
Yes, but the market has matured and entry-level competition is far stronger than it was several years ago. Both halves of that sentence are true simultaneously, which is why the question keeps getting answered badly in each direction.
The Bureau of Labor Statistics still projects faster-than-average growth for the occupation, so this is not a shrinking field. What has changed is who captures the premium within it.
That premium has shifted towards candidates who combine statistical skill with engineering ability — people who can build the thing and not only analyse it. Purely analytical profiles now compete against a large supply of trained graduates, which is exactly where the entry-level pressure comes from. If you are deciding what to learn next, the engineering half is where the scarcity currently sits.
Do you need a master's degree to earn a good data scientist salary?
No. A master's is a common route in and genuinely helps with initial screening, but it is not a requirement at many employers, and treating it as one delays people unnecessarily.
Candidates are routinely hired without one on three profiles: a strong portfolio, relevant domain experience, or an analytics background extended into modelling. That third route is the most under-used and often the fastest, because it builds on work you are already trusted to do.
The degree matters most in two specific places — research-oriented positions, where the training is load-bearing rather than decorative, and employers with formal educational requirements, where it is a filter regardless of ability. Decide by whether your target roles fall into either category, not by whether a degree would generally be nice to have.
Why is my data scientist salary lower than figures I see online?
Because those figures usually describe a different job market than yours. Online numbers typically reflect total compensation at large technology employers in high-cost cities, which is not the median experience of the occupation by a wide margin.
Role weighting is the other half, and it is the part people miss. Reporting-heavy data science pays less than engineering-heavy data science carrying the identical job title — so two people can both be "data scientists" and be doing work the market values very differently.
Compare against official occupational data and local job postings rather than viral compensation screenshots. Postings for your city and your seniority are the most honest benchmark available, and they have the advantage of describing jobs you could actually apply for rather than an unrepresentative tail.
Does specializing in a domain increase data scientist pay?
In some fields substantially, in others barely — and knowing which is which prevents a lot of misdirected effort.
Domain specialisation increases pay where mistakes are costly and regulation is heavy: healthcare, insurance, risk and pricing. Employers there value someone who understands the subject matter as well as the statistics, because that combination is genuinely scarce and because the cost of a modeller who does not understand the domain is measured in something other than accuracy.
In general technology roles it matters much less than engineering depth. If you work in consumer software, another year of domain knowledge will do less for you than learning to deploy and monitor what you build. The rule of thumb: specialise into domains where being wrong is expensive, and into engineering everywhere else.
How can a data analyst move into a data scientist salary band?
Add engineering and modelling work to your current analyst role rather than starting over. The restart is the instinct and it is almost always the slower path.
Concretely, three moves in order: automate a reporting pipeline so it runs without you, build a predictive model that someone actually uses in a decision, and learn enough software engineering practice to deploy and maintain it. Each of those is a thing you can point at, and the third is what most analysts are missing.
Internal moves are usually easier than external ones, which surprises people who assume they need a new employer to get a new title. Your current employer already trusts your judgement about their data — that trust is an asset you would be throwing away, and rebuilding it elsewhere takes longer than acquiring the missing skills where you are.
Is remote data science work paid less?
It depends entirely on the employer's compensation policy, and there is no industry norm to assume in either direction.
Companies using location-adjusted bands pay remote employees according to where they live. Companies with a single national band do not, and pay the same rate regardless of address. Both approaches are common, so this is a question to ask early rather than a risk to price in.
For candidates outside major metropolitan areas, a remote role at a national-band employer is often the highest-paying option available — better than local on-site work by a margin that no skill decision could close in the same timeframe. That makes the banding policy one of the most consequential things to establish about any remote job, and it is usually stated plainly in the offer if you ask.
Keeping this current. Course formats, prices, and certification exam fees change and vary by region. We review our guides regularly, and we always recommend confirming the specifics on the provider's official page before you enrol.