Data Scientist
Data scientists build the statistical models and machine learning systems that forecast demand, sort customers and automate decisions for an organisation.
- Median salary*
- $124,800
3.9%vs last year, before tax
- People employed
- 3,400
3.0%vs last year
- Projected growth*
- +24.6%
to 2035
- AI exposure*
- Moderate
- 6 out of 10
- Average hours*
- 41/wk
+1h vs all jobs
- Shortage status
- Not in shortage
national
- Data scientists work in banks, insurers, government agencies, universities, consultancies and technology companies, usually in a team alongside engineers and analysts.
- The difference from a data analyst's job is that you build predictive models and put them into production, not just report on what has already happened.
- Employers usually expect a degree in statistics, mathematics, computer science or another quantitative field, and some finished work with real data to show for it.
How much do data scientists earn?
The median full-time salary for a data scientist is $124,800 per annum, before tax, up $25,900 since 2018.
Base salaries sit higher in banking, insurance and technology than in universities and government, and private-sector offers often include a bonus or equity that changes the real figure. Contract and consulting work is usually quoted as a day rate, which has to cover your own leave, super and quiet periods, so it is not directly comparable to a permanent salary. If you are weighing up two roles, ask what the bonus is based on and whether it was paid in full last year.
What does a data scientist do day to day?
The list below is what fills most weeks; the exact mix shifts with seniority and whatever stage the current work is at.
- Cleaning and reshaping messy data before any modelling can start
- Chasing down where a number in a dataset came from, often by asking the team that collected it
- Testing a model against real-world outcomes, not only its accuracy score
- Sitting with the people who will use the result and pinning down the question they actually need answered
- Rebuilding a pipeline because a stakeholder wants the data cut a different way
What skills do data scientists need?
Employers look for machine learning, statistical modelling, data analysis, backed by Python fluency and strong problem solving.
Specialist skills
- Machine learning
- Statistical modelling
- Data analysis
Software and tools
- Python
- SQL
- Spark
- cloud ML platforms
General skills
- Problem solving
- Presenting and data storytelling
Is the job growing?
About 3,400 people work as data scientists in Australia, and employment is projected to grow 24.6% over the decade to 2035. That's very strong growth. Few roles in Australia are expanding this fast, and it points to solid demand for years to come.
How do you become a data scientist?
Here's the path most data scientists take, step by step.
- 1Complete a quantitative degree
Statistics, mathematics, computer science, actuarial studies or economics with strong maths all work. The degree usually runs three years full time, with an honours year common for research or specialist roles.
- 2Build a portfolio of finished analysis
Employers want to see completed work: a competition entry, an open dataset you cleaned and modelled, or analysis you did in a current job. Publish the code on GitHub so someone can read it.
- 3Start in a related role and move across
Many people enter as a data analyst, a graduate in a government program or a software engineer, then shift into modelling within a couple of years once they know how the organisation's data fits together.
- 4Add postgraduate study if you are changing fields
A graduate certificate, graduate diploma or master's in data science gives people with a degree in another area the statistics and programming they are missing, and many run part time for people already working.
- 5Learn the production side of the work
SQL, version control and one cloud platform such as AWS, Azure or Google Cloud let you run and update your own model once it is live, without waiting on an engineer.
Study routes we have checked
Qualifications that are a recognised way into this role, and what else you need.
- Bachelor of Business/Bachelor of Data Science
A recognised route into this role
Complete relevant data-science coursework and projects. Data engineering also requires practical database and programming capability; employer requirements vary.
- Bachelor of Computer Science/Master of Data Science
A recognised route into this role
Complete relevant data-science coursework and projects. Data engineering also requires practical database and programming capability; employer requirements vary.
- Bachelor of Data Science
A recognised route into this role
Complete relevant data-science coursework and projects. Data engineering also requires practical database and programming capability; employer requirements vary.
- Bachelor of Mathematics/Master of Data Science
A recognised route into this role
Complete relevant data-science coursework and projects. Data engineering also requires practical database and programming capability; employer requirements vary.
- Bachelor of Science/Bachelor of Data Science
A recognised route into this role
Complete relevant data-science coursework and projects. Data engineering also requires practical database and programming capability; employer requirements vary.
- Bachelor of Science/Master of Data Science
A recognised route into this role
Complete relevant data-science coursework and projects. Data engineering also requires practical database and programming capability; employer requirements vary.
- Graduate Diploma in Data Science
A recognised route into this role
Complete relevant data-science coursework and projects. Data engineering also requires practical database and programming capability; employer requirements vary.
- Master of Artificial Intelligence
A recognised route into this role
Covers the modelling and analysis used in data science. Employers usually also look for experience applying it to real data sets.
- Master of Data Science
A recognised route into this role
A recognised route into data science. Employers usually look for programming skills and project work on real data alongside the degree.
- Master of Business Analytics
A step towards this role
Provides a base in modelling and analysis. Data science roles often expect deeper programming and machine learning skills than some analytics programs cover.
Ready to apply as a data scientist?
Whether you're working toward becoming a data scientist or already are one and want a hand with the next step (sharpening your resume for ATS screening, tightening your cover letter, or knowing what you'll actually be asked at interview), here are examples grounded in this specific role, not generic templates.
What jobs can a data scientist move to?
Moving into Quantitative Analyst typically comes with the biggest pay rise, worth $23,400 a year more on average.
- Quantitative AnalystSuggested move$148,200+$23,400
- Quantitative TraderSuggested move$148,200+$23,400
- Machine Learning EngineerSuggested move$127,400+$2,600
- MathematicianSuggested move$111,800−$13,000
Thick lines are moves people were recorded making in Jobs and Skills Australia's Data on Occupation Mobility (income tax records, 2011-12 to 2020-21). Thin lines are moves we suggest from skill overlap and entry requirements.
| Move to | Typical pay change | Overlap | Retraining |
|---|---|---|---|
Quantitative Analyst Quantitative modelling and programming carry into quantitative analysis, with short course study adding finance and market knowledge. | +$23,400 | 50% | short course |
Quantitative Trader Quantitative modelling and coding skills carry into quantitative trading, with short course study adding market microstructure and risk knowledge. | +$23,400 | 50% | short course |
Machine Learning Engineer Quantitative modelling and coding skills carry into machine learning engineering, with short course study adding production deployment. | +$2,600 | 48% | short course |
Mathematician Mathematical modelling and computational skills carry into mathematician roles, applying theory to practical problems with little further study. | −$13,000 | 63% | minimal |
Statistician Quantitative modelling and computational skills carry into statistician roles, adding deeper theoretical foundations through substantial study. | −$18,200 | 63% | reskill |
Moves are chosen from Jobs and Skills Australia's Data on Occupation Mobility, which follows income tax records between 2011-12 and 2020-21, together with entry requirements and skill overlap. A known move is one people were seen making in that data. Pay change compares median full-time pay for the two roles.
Who works as a data scientist?
The typical data scientist is 34 years old; 73% are men, 88% work full-time, and full-timers average 41 hours a week.
- 34
- Median age
- 27%
- Female share
- 88%
- Full-time
- +1h
- vs all-jobs avg
What's it like being a data scientist?
You will spend more of your week on cleaning data, talking to people and checking results than on the modelling itself, which surprises a lot of newcomers. Deadlines come from the business cycle, so a model needed for a pricing review or a semester's enrolments sets your pace. If you like following a puzzle through to an answer and then explaining it to someone who does not code, this work tends to hold your interest. If you want to spend your days writing models and little else, the meetings and the tidying up will grate.
What people like
- Puzzles that take days to crack. You get a vague request from a business area, and turning it into something measurable means following leads and ruling out explanations until a pattern holds.
- Seeing your work change a decision. A pricing rule, a fraud check or a staffing roster changes because of what your model showed, and the effect turns up in the numbers months later.
- Always something new to learn. Libraries, platforms and methods move quickly, so there is steady room to pick up new tools on the job without leaving the role you are in.
- Being the person who can explain it. When you can take a complicated result and make it clear to people who do not code, you get pulled into decisions early rather than handed the leftovers.
What people find hard
- The data is rarely ready. Records go missing, teams define the same field differently and formats change, so preparation takes longer than the modelling does.
- Requests arrive without a decision attached. Someone asks for a prediction without saying what they would do differently if it came out high or low, and you have to go back and ask.
- Good models that go nowhere. A model can be accurate and still be shelved because nobody has built the system to run it or agreed to change the process it would affect.
- Explaining uncertainty to people who want a number. The honest answer is a range with caveats, and holding that line with a manager who wants one figure takes patience.
Based on our synthesis of professional-body surveys and public accounts of the role, not first-person verified reviews.
Which industries employ data scientists?
Professional, Scientific and Technical Services employs the largest share of data scientists, followed by Financial and Insurance Services.
Top employing industries
- 1Professional, Scientific and Technical Services
- 2Financial and Insurance Services
- 3Public Administration and Safety
- 4Education and Training
- 5Information Media and Telecommunications
Ranked by employment share; the source doesn't publish an exact percentage per industry.
| Bachelor degree | 52% | |
|---|---|---|
| Postgraduate | 26% | |
| Diploma / Advanced Diploma | 13% | |
| Other | 9% |
Will AI replace data scientists?
Moderate exposure. Not replaced, but changed. AI is taking on some tasks and the rest of the job shifts around them.
Automated tools now handle a lot of the standard modelling work: they will clean a dataset, fit a baseline model and draft the code, which is why this role has moderate exposure. What they cannot do is decide whether the data means what it appears to, or whether the model answers a question anyone cares about. For someone entering the field today, the safest ground is production work, cloud platforms and the ability to explain a result to a non-technical audience.
How a typical week splits
- Data cleaning and preparation30% of the week
Sorting out missing values, duplicate records and inconsistent categories still needs someone who understands what the fields are supposed to mean.
- Exploratory analysis and first-pass reporting25% of the week
Tools will produce summary statistics and charts from a prompt, so the value moves to choosing which comparison actually answers the question.
- Model training and tuning20% of the week
Automated platforms will fit and compare standard models, but deciding whether one is good enough for the decision at hand stays with you.
- Framing the problem and explaining results25% of the week
Getting a business area to say what decision they are making, and what would change their mind, is a conversation no tool holds.
How we work this out
We break each job into its main tasks, estimate the share of a typical week each one takes, and judge how far current AI and automation tools reach into it. The score out of 10 starts from the job's overall rating (low is 1 to 3, moderate 4 to 6, high 7 to 10) and the task split places it within that range. Source: Career Tips analysis of each role's main tasks.
Common questions about becoming a data scientist
Straight answers to the questions people ask most.
How much do data scientists earn?
Data scientists earn $124,800 per year before tax at the median for full-time work. Pay climbs with experience and with the scale of the team you join, and a lead or principal data scientist earns well above that figure. Ask what counts as senior in the team you are applying to, because the titles vary between employers.
How do you become a data scientist?
Most start with a degree in statistics, mathematics, computer science or another quantitative field, then build evidence that they can work with real data. A portfolio, whether from open datasets, competitions or analysis done in a current job, is what shows an employer you can finish something. If your degree is in another area, a graduate certificate or master's in data science is a common way across.
Are data scientists in demand?
Data scientists are currently not in shortage, and employment is projected to grow 24.6% over the decade to 2035. Openings cluster in banking, insurance, government and technology, and employers increasingly want people who can get a model into production. Practical experience counts heavily when you are applying, so an internship, a cadetship or a project inside your current job makes the entry easier.
Will AI replace data scientists?
Automated machine learning tools now handle a lot of the standard modelling and code writing, so the tasks that repeat are the exposed ones. What they do not do is judge whether a dataset can be trusted or whether a question is worth answering, and that judgement is what teams keep paying for. Learning the cloud platforms your employer runs, and how to explain a result to people outside the team, protects you more than adding another modelling technique.
What can a data scientist move into?
Machine learning engineering keeps the modelling and adds production deployment, with a pay difference of $2,600 more. Quantitative analysis in finance applies the same maths to markets, and usually needs some short further study in finance and risk. Statisticians and mathematicians also move into this work, adding programming and business context as they go.
Do data scientists code every day?
Most do: Python and SQL are the working tools, and a day without any coding usually means you were in meetings or fixing data by hand. You do not need to be a software engineer, but you do need to read code, debug it and use version control. Cloud platforms and notebooks have lowered the barrier, and the standard is set by the team you join.
Related roles
- Machine Learning Engineer
- Quantitative Analyst
- Quantitative Trader
- Mathematician
- Statistician
- Data Analyst
Not sure this is you? Take the career quiz and get a ranked shortlist of roles that fit how you like to work.
Career advice
- How to change careers, and which moves are realisticData on 1,573 career moves across 350 Australian occupations shows 63.3% need no more than a short course, though skill overlap alone won't tell you which ones. Six worked examples, and how to check your own move.
- Skills to put on your resume (with skills finder)A resume skills section built for the job you're applying for. Search your occupation to see the specialist skills, tools and general skills it calls for, with worked examples and what to write if you have no work history yet.
- How to prepare for a job interviewWhat to work through before a job interview: a prep sheet built from the job ad, the stories that cover most questions, and the questions an interviewer can't lawfully ask in Australia.