Data Methodology
This document is the source-of-truth for how every number on ratebench.app is computed. None of the v1 numbers are from real freelancer submissions yet — that's what the contributor flywheel is being built to fill in. Every claim on this site should be reproducible from a public source via the formula documented below.
1. Sources
1.1 BLS OEWS (Occupational Employment and Wage Statistics)
The base layer is the U.S. Bureau of Labor Statistics OEWS program, May 2024 reference period. BLS publishes P10, P25, P50, P75, and P90 percentile wages for hundreds of occupations across 600+ metropolitan areas.
| SOC code | BLS occupation | Ratebench label | National P50 (hourly) | Source |
|---|---|---|---|---|
15-1252 | Software Developers | React Developer | $63.98 | bls.gov ↗ |
15-1255 | Web and Digital Interface Designers | UX Designer | $47.16 | bls.gov ↗ |
27-1024 | Graphic Designers | Brand Designer | $29.47 | bls.gov ↗ |
27-3043 | Writers and Authors | Copywriter | $34.74 | bls.gov ↗ |
27-4021 | Photographers | Photographer | $20.44 | bls.gov ↗ |
1.2 What's NOT in v1
- MSA × occupation data. BLS publishes city-level breakouts, but v1 uses national figures only. City-level adjustment is on the roadmap.
- Real freelance submissions. Contributor count is N=0 at launch. This will populate as freelancers submit real rates via /contribute.
- Vertical-specific data(e.g., "UX in fintech vs. UX in e-commerce"). Coming with N>500 contributors.
2. Freelance markup ladder
BLS gives W-2 employee hourly wages. Freelancers must charge more to cover benefits, employer contributions, business overhead, profit, and risk premium. The Ratebench markup ladder translates BLS percentiles to estimated freelance rates:
| Percentile | Multiplier over BLS | Rationale |
|---|---|---|
| P25 (entry / discount work) | × 1.4 | Entry-level freelance, low leverage, often agency-mediated |
| P50 (median freelance) | × 1.7 | Standard markup: ~30% benefits + ~8% employer payroll tax + ~25% overhead + modest profit |
| P75 (proven freelancers) | × 2.1 | Established freelancers with reputation, repeat clients, niche expertise |
| P90 (specialists with leverage) | × 2.7 | Top-tier specialists, sole-source contracts, scarcity premium |
The ladder increases with percentile because pricing power scales with experience, reputation, and specialization. Industry guidelines from Robert Half, Toptal, and Codementor consistently put freelance markup between 1.5x (junior) and 3x (specialist). Our ladder sits inside that range and is conservative on both ends.
3. Tier mapping (Junior / Mid / Senior)
BLS doesn't publish wages by experience level. Ratebench's tier mapping is a heuristic that adjusts the base BLS median by an experience multiplier:
| Tier | Base = BLS P50 × | Rationale |
|---|---|---|
| Junior | 0.7 | Roughly the BLS P15–P25 slice — early-career workers |
| Mid | 1.0 | The BLS median — workers in their 5–10 year band |
| Senior | 1.4 | Roughly the BLS P75–P85 slice — experienced practitioners |
Once contributor submissions reach N≥30 per (skill × tier), tier-specific medians from real reports will replace the heuristic.
4. Worked example: UX Designer, Mid tier
BLS OEWS 15-1255 (Web and Digital Interface Designers)
National median hourly wage, May 2024: $47.16
Step 1 — tier-adjusted base:
Mid tier multiplier = 1.0
Base = $47.16 × 1.0 = $47.16
Step 2 — freelance markup at P50:
P50 markup = 1.7
Freelance P50 = $47.16 × 1.7 = $80.17/hr
Ratebench shows: Mid UX Designer, freelance median = $80/hr
5. Honest limitations of v1
- No real freelancer submissions yet. N=0 for every cell at launch. The contributor flywheel is the product.
- National figures only. City premiums (SF, NYC) are real — but not modeled until MSA × occupation data lands.
- Tier mapping is heuristic, not measured. Junior/Mid/Senior cutoffs will be revised once real data arrives.
- BLS data excludes self-employed workers by design. The markup ladder compensates, but it's an estimate, not a measurement.
6. Roadmap to better data
| Milestone | What it unlocks |
|---|---|
| First 50 contributors | First real-data sanity check; methodology validation; tier definitions get refined |
| 200 contributors | Tier-specific medians replace heuristic for biggest skill + tier cells |
| 500 contributors | First MSA × occupation breakouts (top 10 cities) |
| 1,000 contributors | Quarterly Bench issue with trendlines; Pro tier data integrity |
Help us replace BLS-anchored estimates with real data.
Contribute your rate →Citations
All BLS data: U.S. Bureau of Labor Statistics, Occupational Employment and Wage Statistics, May 2024 estimates.