Skip to content

Ghost job report

Half the advice is about writing a better CV. Nobody checks whether the job is real.

We re-read every listing in our index each night and record the first and last time we saw it. That tells us something a job board never shows you: how long a role has been continuously advertised. Across 81,708 live listings from 20 sources and 20,600 companies, the median role is filled and gone in 9 days.

26%
still advertised after 30 days
14.6%
still advertised after 45 days
9d
median time a listing lasts

That means 21,274 roles in our index right now have been open for over a month — more than 3.3× as long as the typical role takes to fill. Some of those are genuinely hard to hire for. Others were filled long ago, or were never a real opening: posted to build a CV pile, to satisfy a policy, or to look like a company that is growing.

You cannot tell which is which from the posting. That is the point — and it is why applying faster to more of them is not the win the rest of this industry is selling.

How we measure it

  • Continuous listing, not reposting. A role counts from the first night we saw it to the most recent, matched on a normalised fingerprint of title, company and city — so the same job appearing on four boards is one role, not four.
  • These are floors, not true rates. We have been watching since 2026-06-07 — 160 days. A listing that went up before then looks younger to us than it really is, so the real figures are higher and we cannot yet say by how much.
  • We report duration, not intent. A long-running listing is a fact. Calling it a ghost job is an inference, and we do not name companies.

Spend your applications on the roles that are real

Ryser flags listings with ghost-job risk and hides them by default, then tailors your CV to the ones worth your time — grounded in your real experience, never invented.

Figures recomputed nightly from the live index, after that night’s intake and prune have settled. Page rendered 2026-11-14. Methodology above.