Tool 1 of the Get Paid Kit · For data scientists

Data Scientist hourly rate calculator

This free calculator works out the hourly rate a freelance data scientist should charge, starting from the income you need — not from a day rate quoted on a forum. It counts the data cleaning, the stakeholder questions and the ‘one quick look’ requests that surround the modelling itself, and divides by the hours you can actually bill. With the honest presets below it lands near $158 an hour: a twenty-hour diagnostic sprint prices at $3,160, and a sixty-hour model build at $9,480.

Runs in your browser. Nothing you enter is uploaded, sent to a server or stored anywhere except this device.

The pricing reality for freelance data scientists

Freelance data science sells answers a business cannot staff for: churn models, forecasts, pricing analysis, experiments. The clients have data — usually messier than they think — and a question worth money. Your rate has to carry both halves of that work: the modelling they asked for, and the cleaning they did not know they were buying.

Keep the neighbouring trade separate. A data analyst reports on what happened — dashboards, BI and recurring reporting — and this site prices that work on its own page. A data scientist builds what predicts or decides: the model, the forecast, the experiment. Clients mix the two titles up constantly, and the mix-up usually arrives as model work priced like reporting. The distinction is worth one calm paragraph in every proposal.

The preset below starts billable time at 55%. Analysis work grows an unbilled tail: scoping conversations that go nowhere, data-access wrangling, the ‘quick look’ that becomes an afternoon. The numbers are starting presets, not a survey: a $110,000 target income, $3,800 a year of expenses — software, the cloud compute for jobs your laptop cannot hold, and the courses that keep the methods current — and 55% billable time. With those starting numbers the calculator lands near $158 an hour.

Your numbers — data scientist presets

Desired annual income ($) what you want to take home before tax

Annual business expenses ($) software, cloud compute and courses

Tax estimate (%) — planning estimate, not tax advice 25%

Vacation weeks per year

Hours per week

Billable time (%) 55% modelling, analysis and dashboards bill; data wrangling overruns, pitching and admin do not — 55% is a realistic start

Profit margin (%) a buffer for quiet months and savings

Your recommended rate

$0 / hour

$0Day rate (8 hours)

0Billable hours per year

Income + expenses needed	$0
Grossed up for tax	$0
With profit margin	$0
Working weeks	0
Total working hours	0
Billable hours	0

Your rate is saved on this device only, so the Proposal Generator can show it when you price a project.

Easy to miss

Costs data scientists forget to put in the rate

The cleaning nobody budgets for. Client data arrives with duplicates, gaps and three spellings of the same customer. The hours spent making it analysable are real production time; a rate that ignores them donates them.
Compute and data access. Cloud compute for the jobs your laptop cannot hold, notebook and tooling subscriptions, the occasional dataset. Small lines, every month.
The ‘one quick look’. Stakeholders ask for quick checks between deliverables. Each one is small; together they are a day a month of unbilled analysis if the rate does not carry them.
Methods move. Libraries, techniques and platforms change every year. Courses and reading time keep a data scientist employable — and priced like it.
Experiments that find nothing. Some honest tests end with ‘the data cannot answer that’. You still did the work; the rate and the scope both have to survive negative results.

The maths

How this calculator works

Start with the hours you can genuinely sell. Take 52 weeks, subtract your vacation weeks, multiply by your hours per week, then multiply by your billable percentage. With the data scientist presets — four weeks off, 40 hours a week, 55% billable — that is 1,056 billable hours a year: the modelling, analysis and reporting hours left once data-wrangling overruns, stakeholder questions and pitching have taken theirs.

Next, work out what those hours must bring in. Add your desired income and your annual business expenses together — $113,800 on the presets — gross the total up for tax by dividing by one minus your tax rate — at a 25% estimate, divide by 0.75 to get $151,733 — then add your profit margin to reach $166,907 and divide by your billable hours. That final number, $158, is the rate below which the year stops adding up.

Now convert it into the units clients buy. A full day at the preset rate is $1,264. A focused diagnostic sprint — twenty hours inside a client’s data, ending in a written answer — is $3,160. A model build scoped at sixty hours is $9,480. Every fixed price you quote should survive this multiplication; if it cannot, the scope, not the rate, is what changes.

Disclaimer: This is a planning estimate, not tax advice. Tax rates, deductions and allowances vary by country, state and business structure. Enter the rate you expect to pay, and check significant decisions with a qualified accountant.

Next: build a proposal with this rate Use the general rate calculator

Questions

Data Scientist rate questions

How is a freelance data scientist’s work different from a data analyst’s?+

Analysts report on what happened — dashboards, BI and recurring reporting. Data scientists build what predicts or decides: models, forecasts and experiments. The work overlaps at the edges and some contracts blend both, but the model-building end carries scarcer skills and longer, less predictable tasks, which is why these presets start from a higher target income than the analyst page. If a client asks for ‘a dashboard with some AI in it’, price the parts separately.

Should data scientists charge hourly or per project?+

Hourly or daily for open-ended analysis and embedded contract work, where the questions change as the data answers back. A fixed project price works when the deliverable is bounded — one model, one forecast, one experiment — and it should still be converted from your hour, with the data-cleaning risk priced in rather than discovered. Discovery work, where nobody yet knows whether the data can answer the question at all, belongs in a paid first phase.

What if the analysis finds nothing — do I still charge?+

Yes. A rigorous test that rules something out is a result, and it was the client’s question, not your promise of a particular answer. This is why the scope has to sell the investigation, not the outcome: ‘test whether churn can be predicted from this data’ is honest work; ‘a model that predicts churn’ is a promise nobody can make before seeing the data. Price the investigation, deliver the finding.

Rates differ by trade — the same maths runs with different presets for graphic designer rates, freelance writer rates, web developer rates, marketing consultant rates, photographer rates, video editor rates, virtual assistant rates, social media manager rates, bookkeeper rates, translator rates, business coach rates, illustrator rates, copywriter rates, UX designer rates, SEO consultant rates, data analyst rates, motion designer rates, technical writer rates, podcast editor rates, email marketer rates, life coach rates, proofreader rates, event planner rates, voice actor rates, online tutor rates, personal trainer rates, interior designer rates, music producer rates, grant writer rates, UX researcher rates, book editor rates, transcriptionist rates, presentation designer rates, brand strategist rates, video producer rates, art director rates, WordPress developer rates, mobile app developer rates, 3D artist rates, game developer rates, Shopify developer rates, online business manager rates, podcast producer rates, data entry specialist rates, UX writer rates, copy editor rates, wedding photographer rates, resume writer rates, instructional designer rates, business analyst rates and email designer rates, or start from the general freelance hourly rate calculator.