Job Counts vs. Job Postings: How BLS Measures Where the Jobs Actually Are

Most “most in-demand jobs” and “hottest job markets” lists are built by scraping job postings from boards like Indeed and LinkedIn. Posting counts are easy to collect and update daily, which is why they dominate headlines — but they measure something quite different from how many people actually hold a job.

What a posting count actually measures

A job posting is a recruiting event, not a job. Posting volume rises with turnover (occupations that churn workers post constantly, whether or not they're growing), with recruiting practice (some employers repost the same role weekly or post “evergreen” ads with no specific opening behind them), and with duplication across boards. An occupation with high turnover and aggressive recruiting can generate ten times the postings of a larger, more stable occupation. Postings tell you where hiring activity is loud; they don't tell you where the jobs are.

What an employment count measures

The BLS Occupational Employment and Wage Statistics (OEWS) program publishes, for each occupation in each metro area, an estimate of total employment — the number of filled wage-and-salary jobs on employer payrolls, reported by the employers themselves in a survey covering about 1.1 million business locations over a three-year cycle. The current release on this site is the May 2025 reference period. These counts move slowly, because actual employment moves slowly — and that stability is the point. When the most common jobs ranker says an occupation employs more people in a metro than any other, that's a published federal estimate of filled jobs, not an echo of recruiting volume.

Raw counts favor big metros — read the share column

New York and Los Angeles will top almost any raw-count ranking simply because they contain the most jobs of every kind. The honest correction is to divide two published figures: an occupation's employment in an area divided by that area's total published employment. A share of 3% of all local jobs says something a raw count of 50,000 doesn't — it tells you the occupation is genuinely concentrated there rather than just present in a large market. BLS publishes a formal concentration measure (the location quotient) that this site doesn't currently ingest; the share column is the same idea computed transparently from two published numbers.

When BLS won't publish a number, neither do we

For some occupation-and-area combinations, the sample is too small to meet BLS publication standards, and the employment estimate is suppressed. A ranked list has exactly one honest way to handle that: exclude the row. Estimating a stand-in value or substituting a broader geography inside a ranking would silently mix data of different reliability — so suppressed rows simply don't appear.

What employment counts can't tell you

A large employment count means many people hold the job today. It does not mean the occupation is growing, hiring, or easy to break into — a huge, stable occupation can have few openings, and a small one can be expanding fast. Openings and growth are measured by different BLS programs (JOLTS and the Employment Projections), which are outside this tool's scope. Use employment counts for what they are: a map of where jobs exist. For what those jobs pay, the highest-paying jobs ranker runs the same two ranking directions over median wages instead, and the percentile lookup gives the full pay range for any occupation-and-metro pair you find.

Informational only, not professional or financial advice.