Data Sources
CareerOS uses six primary data sources to produce salary estimates. Each source has different strengths and limitations — we triangulate across them to produce more accurate ranges than any single source provides.
U.S. Bureau of Labor Statistics (BLS)
The BLS Occupational Employment and Wage Statistics (OEWS) program conducts a semi-annual mail survey of approximately 1.1 million non-farm business establishments. It is the most comprehensive occupational wage dataset in the US and serves as our baseline anchor for all salary estimates.
Strengths
Nationally representative, methodologically rigorous, covers all industries
Limitations
Lags market by 12–18 months; excludes self-employed; may undercount equity-heavy tech compensation
LinkedIn Salary Insights
LinkedIn aggregates salary data from users who voluntarily submit their compensation information. With 900M+ members globally, it represents the largest self-reported compensation dataset. We use LinkedIn Salary to validate BLS figures and identify recent market movements.
Strengths
Current market data, role-specific, filters by experience/industry
Limitations
Self-selection bias (higher earners more likely to report), US-primary, excludes total comp
Glassdoor Compensation Reports
Glassdoor requires users to share salary data to access peer salary information, creating a more complete participation model. We use Glassdoor data primarily for city-level validation and company-specific compensation patterns.
Strengths
Company-specific data, city-level filters, includes bonus information
Limitations
Self-selection bias, company averages vary widely by company culture
Levels.fyi
Levels.fyi focuses specifically on technology companies and requires users to submit verifiable compensation packages including base salary, equity (RSUs/options), signing bonus, and level/title. This provides the most accurate total compensation data for technology roles.
Strengths
Verified data, includes total comp breakdown, company and level specific
Limitations
Tech-only, skewed toward senior roles and large companies
Job Posting Salary Data
Since 2023, California, New York, Colorado, and Washington require employers to disclose salary ranges in job postings. This mandated disclosure provides current market data directly from employers. We analyze posting data to validate our estimates for these states.
Strengths
Current market rates, employer-disclosed, reflects actual job offers
Limitations
Only available in transparency-law states, ranges can be very wide
Council for Community and Economic Research (C2ER)
The C2ER Cost of Living Index measures the relative price levels for consumer goods and services by city. We use C2ER data to calculate city-level cost of living adjustments and purchasing power comparisons in our comparison tools.
Strengths
Widely used benchmark, consistent methodology across cities
Limitations
Measures consumer goods, not always specific to professional household budgets
How We Combine These Sources
For each role and location, we start with the BLS baseline, then compare it to the self-reported median from LinkedIn and Glassdoor. When the three sources agree within 15%, we use the median value. When they diverge significantly, we investigate the discrepancy — usually because the role definition differs or because one dataset has a sampling issue for that geography.
For tech roles specifically, BLS often understates current market rates because it lags by 12–18 months. In those cases, we weight the Levels.fyi and LinkedIn data more heavily and note the discrepancy in our methodology.