Methodology

How the data works.

Every number on this site comes from public USAJOBS data and can be traced back to its source. Here is where it comes from and how each score is calculated.

Where the data comes from

Job announcements: the USAJOBS API. This build covers 244,223 announcements opened from December 1, 2025 to November 15, 2026, plus 12,260 announcements open on November 15, 2026.
Series titles and job families: the official occupational series list from USAJOBS.
Series descriptions: summaries of the OPM Handbook of Occupational Groups and Families.

What gets counted

A series is listed when it had at least one job announcement in the last 12 months.
Canceled announcements are not counted.
An announcement that lists several series counts once for each of them.
The data is refreshed every week.

The scores

Demand score (0 to 100): How a series' number of postings in the last 12 months ranks against every other series. A score of 90 means it had more postings than about 90 percent of series.
Remote opportunity score (0 to 100): The share of the series' jobs open right now that are telework-eligible or fully remote. Shown as "Not enough data" when fewer than 5 jobs are open.
Veteran opportunity score (0 to 100): The share of the series' postings in the last 12 months that veterans can apply to: those open to the public or listing a veterans hiring path.
Trend: Postings in the last three complete months compared with the three months before: Rising (up 10 percent or more), Falling (down 10 percent or more), or Steady.

Hiring patterns

Series pages with enough history show how postings usually rise and fall through the year and how many were posted each federal fiscal year. These use every announcement opened from Mar 2017 to Oct 2026, not just the last 12 months.

Busiest and quietest months: each month's postings are compared with the average for the surrounding year, and the typical result for each calendar month is used. Busiest months are named only when the peak is at least 15 percent above an average month.
Smaller series: a series with few postings has a noisy pattern, so its pattern is blended with the government-wide pattern. The more postings a series has, the more its own pattern counts. Series averaging fewer than 5 postings a month show no monthly pattern.
Fiscal years: only complete fiscal years (October 1 to September 30) are shown.

The outlook

We built a three-month outlook of postings for series averaging at least 10 postings a month, and tested it month by month using only the data that was available at the time. Our standard: the actual number should land inside the stated range at least 72 percent of the time, and the typical miss should be 30 percent or less.

2019 to 2024: it met the standard. Actual postings landed inside the range 79% of the time, with a typical miss of 12%.
2025 and 2026, during the hiring freeze: it did not. Actual postings landed inside the range 60% of the time, with a typical miss of 31%.

So the site does not show an outlook yet. Each week we test it on new forecasts, and it will appear only after forecasts made since Oct 2026 meet the standard.

An outlook projects job postings, not openings or hires. Hiring freezes, budgets, and new policies can change postings quickly, and the outlook only reflects them after they show up in the data.

Why there is no competition score

USAJOBS does not publish how many people apply for each job, so there is no honest way to measure competition. Instead, each series shows who its jobs were open to: the public, or only federal employees and other special groups.

Limits of this data

These figures count job announcements, not hires. One announcement can fill several positions or none.
Not every federal job is posted on USAJOBS.
Scores are not hiring guarantees and do not say whether you qualify for a job.

Always read the official announcement and its qualification requirements before applying.