How to Tell a Live Opening From One That Won't Hire

What actually causes a posting to sit open with no hire coming?

A handful of ordinary reasons account for most of it, and none of them require anyone to have lied. An evergreen posting stays live because the role hires continuously, retail, hospitality, and high-turnover technical roles are common examples, and treating it as one opening rather than a standing door misreads what it is.

A rehire pool keeps a posting or a folder of finalists open against a role that will likely reopen; a budget that hadn't cleared when the posting went up can freeze a search mid-process; and plenty of postings simply outlive the person who was supposed to close them once the role filled, or keep displaying on an aggregator after the requisition stopped accepting applications. None of these need a villain. They need someone to notice the posting is stale, which frequently nobody does.

How common the underlying pattern is depends entirely on who's counting and what they count as a ghost job. One academic attempt scraped 269,347 Glassdoor interview reviews and, using a language-model classifier, flagged up to 21 percent of those reviews as indicative of a ghost job, a share of self-selected reviews rather than a count of postings ¹. The same paper ran a second, simpler classifier over the identical data and got 1.6 percent, a thirteenfold spread depending only on which method was used ². Greenhouse, whose own platform hosts postings for thousands of employers, separately states that 18 to 22 percent of jobs posted on it in any given quarter get classified as ghost jobs, with no published method behind the number ³. Two independent attempts landing near a fifth is worth taking seriously as a real phenomenon; neither can tell you which specific posting in front of you is the one that fits.

Check these before spending an hour on a posting

A few minutes of checking beats guessing, because every one of the useful tells is visible from outside the company, on the posting itself or one search away from it. None of them proves a posting is a ghost job on its own; taken together, two or three failed checks is a real reason to spend less time than you would on a posting that clears all of them.

• Repost date against the original. A posting refreshed repeatedly with no visible progress, same requisition, new date, is a stronger tell than one simply sitting open.
• Careers site versus aggregator only. A role that exists on the company's own site is more likely actively worked than one you can only find on a board.
• The same requisition ID reappearing. A role that closes and reopens under the same number within weeks often reflects a search that stalled or restarted, not a fabrication.
• A description written for a category, not a person. Generic responsibilities with no specific team, tool, or project named can mean the posting was assembled early and never refined against a real opening.
• A hiring freeze on the record. Public reporting, an earnings call, a layoff announcement, any of these can explain a posting that looks live but isn't being worked.

Some of what looks like a fresh opening is also simply a rehire pool being reactivated rather than a new search starting from zero. Employers doing this are told plainly that interview evidence from an old finalist pool does not carry forward unchanged, because the tools and expectations around a role shift underneath a stale file. A posting resurfacing from that kind of pool can look brand new from outside while running on a much older process internally, which is one more reason the repost date alone tells you less than it seems to.

What does the check actually buy you?

The checks buy a rough triage rather than certainty, and clearing them is no reason to skip applying to anything that still looks ambiguous: postings that clear most of the checks are worth a real hour, and postings that fail several are worth a five-minute application at most, sent as-is with no rewriting.

Legislators are starting to formalize a version of this same distinction. A New York bill that passed both chambers of the state legislature in 2026 would require employers to label a posting in bold text as either a current vacancy with an intended fill date or an explicit statement that it's collecting resumes for future openings ⁴. It is not law yet, and it binds only New York employers if signed, but it names the exact language this article is asking you to look for on your own, and a posting honest enough to carry it voluntarily today is worth noticing.

A stale posting is not a scam you fell for. It's a reason to move the hour somewhere a human is actually waiting on the other end, which is the only thing any of these checks were ever really measuring. If the silence continues across postings that clear every check, that's a different and separate question worth its own read: why hundreds of applications can produce no replies at all covers what the volume itself is doing, apart from whether any single posting was real.

Ask directly when the checks don't settle it

Asking costs less than another round of guessing. If a posting clears some checks and fails others, a short message to a recruiter, or a question in a first-round call, is a reasonable way to close the gap: is this req actively being worked, and is there a timeline for a decision. Most recruiters answer this plainly, because a candidate who asks is easier to keep informed than one who quietly disappears.

A few situations where asking beats waiting:

• The posting has been live for months with no close date. A direct question resolves in a minute what a search engine cannot.
• You've been asked to a screen but the requisition still looks stale online. That's usually reassuring rather than contradictory: a listing can lag behind an active process by weeks.
• A recruiter goes quiet after an initial reply. That is itself information, closer to a real answer than any check run on the posting alone.

"I wanted to check this role is still actively hiring before I put together a tailored application" is a normal, low-stakes question to ask, and a recruiter's answer, or their silence, tells you more in five minutes than any classifier built on public review data ever could.

See a sample report >