"Ghost job" is a term applicants use for a posting an employer isn't actively trying to fill — kept up to collect résumés, to signal growth, or left live after a role was frozen. It describes an intention, and intentions can't be observed from outside a company. What can be observed is how long a posting has been listed.

Can you tell if a job posting is a ghost job?

Not with certainty, and anyone claiming otherwise is inferring. A posting listed for many months and a genuinely hard-to-fill senior role look identical from a careers page. Age is a signal worth weighing, not a verdict.

How many job postings have been listed for a long time?

Of 524,322 live postings we track across 1,077 employers, 99,913 (19.1%) have been listed 90 days or longer, and 23,122 (4.4%) for a year or longer.

Is a long-listed job posting always a ghost job?

No. Roles stay open for many ordinary reasons: a small candidate pool, a slow approval chain, an evergreen requisition an employer genuinely hires against all year, or a listing nobody remembered to take down. Duration is what we measure; it is not the same as intent.

What is a ghost job?

A ghost job is what applicants call a posting an employer isn’t actively trying to fill — kept live to gather résumés, to signal growth, or simply left up after a role was frozen. The term describes an intention, and intentions cannot be read off a careers page.

What can be measured is duration: 19.1% of the 524,322 postings we track have been listed 90 days or longer, and the median has been up 32 days. We publish the duration and stop there, because the step from “listed a long time” to “never meant to be filled” is an accusation, not an inference.

524,322 live postings1,077 employers we crawl dailyas of September 8, 2026How we measure

Where the term comes from

The phrase spread among job seekers to name a specific frustration: applying to something, hearing nothing, and later suspecting the role was never open. The suspected explanations vary — building a candidate pipeline, projecting growth to customers or investors, satisfying a policy that a role be advertised, or plain administrative neglect.

All of those are claims about why an employer did something. We have no access to that, and we are not willing to publish a guess about a named company’s motives. What we do instead is publish the observable part, with its denominator attached, and let you weigh it.

What the duration data shows

32 days

Median age of the 524,322 live postings we track

19.1%

Listed 90 days or longer (99,913 postings)

4.4%

Listed a year or longer (23,122 postings)

The tail is the part people are reacting to when they reach for the word. Most postings are recent and behave the way you would expect. A persistent minority does not: at any given moment 99,913 listings have been up for three months or more, and the oldest we currently track has been listed 5,010 days.

The removal side adds something the age data alone can’t. Of 267,708 postings we watched come down between July 24, 2026 and September 8, 2026; no fleet read on August 23, 2026, August 31, 2026; 4 more days not confirmed complete (see Methodology), 19,136 later went back up under the same listing. A listing that cycles is a real pattern; it still isn’t evidence of intent.

How to use age without over-reading it

Treat posting age the way you would treat any single data point: as one input. A two-week-old posting tells you the employer was doing something recently. A ten-month-old posting tells you the listing has survived ten months, which is worth knowing before you spend an evening on a cover letter — and is not proof of anything.

You can check a specific posting below, see the longest-open postings we track, or read exactly how we date and count them.

How long has that posting been up?

What this doesn’t tell you

We cannot tell you whether any posting is a ghost job. We measure how long a listing has been live and when it comes down. Nothing in that distinguishes a posting nobody intends to fill from a role that is genuinely hard to hire for.

Long-listed is not a synonym for fake. Specialist roles, roles gated behind a slow approval chain, and requisitions an employer really does hire against continuously all produce long-lived postings for entirely ordinary reasons.

Postings we can identify as continuously-hiring are already excluded. Where a listing is clearly an always-open requisition rather than a specific vacancy, we leave it out of these figures rather than counting it as a stale posting. That exclusion is imperfect and it is applied before the numbers above, not after.