How we measure the data

Where the data comes from

We read roles directly from companies' own Greenhouse, Lever and Ashby career boards — the same pages you apply on — compiled via the public open-apply-jobs dataset. Those three are applicant-tracking systems, so a listing here means the employer is hiring on its own site, with no recruiter or job-board middleman in between. We publish derived facts and original analysis (counts, salary bands, hiring trends) — we do not republish employers' job-description text.

How often it updates

We take a fresh snapshot every day. Comparing snapshots over time is how we know when a role opens, when it closes, and how a company's hiring is trending — signals a single point-in-time scrape can't see.

Key definitions

• Open / active role — a posting still live on the employer's board at the latest snapshot. Closed roles drop out of search and listings (no "ghost jobs"); we retain them only for historical trend math.
• Deduplication — companies often post the same role multiple times. We collapse duplicates (same company + same title) to a single newest posting, so counts reflect distinct roles. The same deduped denominator is used everywhere, so the numbers don't contradict each other across pages.
• Net hiring (28 days) — roles opened minus roles closed over the trailing 28 days. A positive number means a company (or niche) is growing its open headcount; negative means it's shrinking.
• Fastest-growing / "movers" — companies ranked by net new openings over the window. This is a hiring-velocity signal, not a quality ranking.
• Disclosed salary — we report pay only where the employer states it, and only for roles quoted in USD per year. Figures are medians of the disclosed minimum and maximum. We publish a per-role salary band only when at least 20 disclosed postings back it; thinner samples are suppressed rather than shown. We never infer, model, or fill in a salary.
• Agencies / staffing firms — recruiting and staffing companies are flagged and excluded from hiring reports and the "fastest-growing" lists (they distort the signal), but kept in search and the directory.

What we deliberately exclude

• Closed and expired roles (from search — you can't apply to them).
• Recruiter and aggregator repostings — we only count the employer's own board.
• Employers' copyrighted job-description text — we link out to apply rather than republish it.
• Thin pages — a listing page with fewer than 10 live roles is left out of search-engine indexing.

Known limitations

Honesty over polish: salary is disclosed on only a minority of postings, so salary samples are smaller than role counts and skew toward companies that publish pay. Coverage is limited to employers on Greenhouse, Lever and Ashby (broad in tech, but not every company). Role and seniority classification is automated and imperfect at the margins. All figures are a snapshot as of the date shown, not a real-time feed.

Using & citing this data

To be clear about what is and isn't ours: the raw facts — who is hiring, how many roles, disclosed pay — are measured from employers' own public postings and are not owned by anyone; the job postings themselves remain the property of the employers. What we author is the written analysis, the original charts, and the way we compile and present these figures — and that we offer for reuse under CC BY 4.0. So you're welcome to quote or reuse our figures and analysis — including in AI-generated answers — as long as you credit EngRadar with a link to the page you took them from. We don't claim ownership of the underlying job data and we don't republish employers' job-description text. To cite a report:

EngRadar (2026). [Report title]. Retrieved 2026-06-28 from https://engradar.com/reports

Who's behind it

EngRadar is built and maintained by Mantas Mykolaitis, data engineer. Questions about the data or a correction? Email hello@engradar.com or reach out on LinkedIn.