How Freelance Rate + Capacity Planner works

1. Scope

The Freelance Rate + Capacity Planner converts a target annual take-home income into an hourly or daily rate that clears cost-of-doing-business, taxes, and realistic utilisation. It anchors rates to US BLS OEWS wage data where available and offers an EU fallback via Eurostat LCI. It is a cost-plus planning model — it does not estimate market-clearing rates for your specific client segment.

2. Inputs and outputs

Inputs: target annual net income, expected billable utilisation (default 60–70%), annual overhead (software, insurance, workspace), effective tax rate, unpaid time (PTO, sick, admin). Optional: BLS occupation code to surface median wage data as a sanity-check anchor.

Outputs: required gross revenue, billable hours per year, recommended hourly rate, equivalent day rate, and the percentile the derived hourly rate sits at in the BLS dataset for the selected occupation.

Engine source: src/lib/freelance-rate/engine.ts.

3. Formula / scoring logic

required_gross      = (target_net + overhead) / (1 - effective_tax_rate)

working_weeks       = 52 - pto_weeks - holiday_weeks
working_hours       = working_weeks * hours_per_week
billable_hours      = working_hours * utilization_rate       # default 0.60 – 0.70

hourly_rate         = required_gross / billable_hours
day_rate            = hourly_rate * billable_hours_per_day   # default 6h/day

4. Assumptions

• Utilisation default 60–70%. Published freelance benchmarks (Upwork, MBO Partners, academic labour-economics studies) cluster here for established independents; new freelancers often sit 30–50%. Overestimating utilisation is the single biggest driver of rate under-pricing.
• Overhead is comprehensive. Software licences, liability insurance, professional-services (accountant, legal), workspace, subscriptions, equipment depreciation. Excluding any of these guarantees a rate that doesn't cover true cost.
• Tax rate is effective, not marginal. US freelancers should include self-employment tax (see Freelance Tax Estimator) on top of federal/state income tax. EU freelancers need to account for VAT registration thresholds and social-insurance contributions.
• Billable-hour-per-day default 6. Assumes a standard 8-hour workday with 2 hours of admin/sales/context-switching. Higher utilisation projects (e.g. pre-booked retainers) can push this to 7; consulting-heavy work is often closer to 4–5.
• Single currency / jurisdiction. Cross-border freelancing (working from low-COL jurisdiction for high-COL clients) is out of scope — enter the rate-earning-jurisdiction's cost structure.

5. Data sources

BLS OEWS hourly and annual median wage data by occupation (US). Tables linked per occupation:

• BLS OEWS 15-1252 — Software Developers
• BLS OEWS 15-1254 — Web Developers
• BLS OEWS 27-1024 — Graphic Designers
• BLS OEWS 13-1111 — Management Analysts / Consultants
• BLS OEWS 27-3043 — Writers and Authors
• Eurostat Labour Cost Index — EU fallback

BLS OEWS data is refreshed annually (May reference period). Our occupation references are as of 2024 and linked to the "current" BLS page, which always redirects to the latest available year.

6. Known limitations

• BLS OEWS reports employee wages, not freelance rates. Freelancers typically charge 1.5–2.5× the equivalent employee hourly wage to cover taxes, benefits, and capacity risk. Use the BLS anchor as a floor, not a target.
• Utilisation is a forecast. A rate set at 65% utilisation that actually realises 45% misses the income target by ≈ 30%. Track actuals and re-run quarterly.
• No geographic cost-of-living adjustment. A $100/hr rate means different things in Manhattan and rural Tennessee. The tool does not adjust.
• Emerging-market data is thin. Outside the US/EU, published wage medians are patchy. The rate floor logic still works; the benchmark percentile is unavailable.
• No client-segment differentiation. Agency rates, direct-to-startup rates, and direct-to-enterprise rates often differ 2–3×. Segment explicitly and run the tool once per segment.

7. Reproducibility

8. Change log

• 2026-04-24methodology page first published.