How these pages are made
Refit's career guides aren't spun from a template with a job title swapped in. Each one is built from real, openly-licensed public data and has to earn its place.
The data
- O*NET 30.3 (US Dept. of Labor) — the tasks, skills, knowledge, tools, and related-occupation graph for ~900 occupations. Used under CC BY 4.0.
- BLS OEWS May 2024 — median and percentile wages and employment, national and by state. Public domain.
- BLS Employment Projections 2024-34 — the 10-year outlook, typical entry education, and annual openings. Public domain.
The 7-gate usefulness check
Before any page is built or added to our sitemap, it has to pass all seven of these — or it's never published:
- At least three distinct, non-empty data blocks for this exact occupation.
- At least one computed insight (a derived analysis, not a copied table).
- A real, scoped resume-tailoring tool for this role.
- A uniqueness check (SimHash) so no two pages are near-duplicates.
- A substance floor — real content well beyond the page chrome.
- Source attribution and a data-version stamp.
- Honest, unique metadata and an honest promise (we never invent your skills).
In the latest build, 15,537 candidate pages were checked; 6,913 passed all seven gates and were published. The rest were rejected — most often for failing the uniqueness or data-presence gate — and never shipped.
The honest tool
Every page lets you paste your real resume and re-angle it toward that role. Refit rewords what you actually wrote to match the role's real requirements — and a no-fabrication gate rejects any change that would invent a skill you didn't claim. If the honest fit isn't there, we tell you what to build instead.