How to Spot a Ghost Job Before You Apply
In a survey of 1,045 hiring managers by Clarify Capital, 43% said they keep listings open for reasons unconnected to filling them, either to keep existing staff feeling replaceable or to make the company look like it's growing. One in ten admitted to leaving a posting live for more than six months.
You are the unpaid audience for that. Every application you send into one of those postings costs you the tailoring time, the follow-up, and the slow erosion of assuming your resume is the problem.
Greenhouse, which runs the hiring for thousands of companies and can therefore count rather than survey, puts it at 18% to 22% of postings in any given quarter, with roughly 70% of the companies on its platform having posted at least one ghost job in a single quarter of 2024. So somewhere near a fifth of what you're looking at isn't attached to a job.
In this post you'll learn why companies run these listings, which seven tools reduce your exposure to them, what none of those tools can actually verify, and the manual checks that catch what's left.
Why the listing is up when the job isn't
Four situations produce most of them, and they're worth telling apart because each leaves a different trace.
Pipeline building is a company gathering resumes for a role it expects to open later, which produces evergreen postings that never close and often say so in the text. These concentrate in high-turnover and early-career hiring, which is why areas like entry-level real estate work carry so many listings that read as permanently open. A req kept warm through a budget freeze looks almost identical from outside, with the difference that the posting was real when it went up and is specific enough to prove it.
The third case is the one that hurts, because a role with an internal candidate already chosen stays posted to satisfy a policy requiring an external search. Everything about it is genuine except the outcome. The fourth is ordinary neglect, a filled role whose posting nobody switched off, which is why the age of a listing carries so much information.
Underneath all four sits the motive the Clarify Capital respondents described, which is posting as advertising. About half said they keep roles open because they're always interested in good people. That sentence is doing a lot of work for the company and none at all for you.
Where the tools help, and where they stop
Start with what a ghost job actually costs, because it changes what's worth paying for. If you're applying by hand, the cost is your evening. If you're using an automated tool like AIApply, where auto-apply runs on credits at roughly 39 cents each on the 100-credit pack, a dead posting has a price on it.
That's the argument for its Review Mode, which holds each application for your approval rather than submitting blind. Thirty seconds of checking against a wasted credit is an easy trade, and the same logic applies to any tool billing you per application.
Your own history is a better detector than most software. Simplify is free, tracks every application you've made, and that dated record is what lets you notice the same role reappearing in your feed six weeks later wearing a fresh posting date. Nothing flags that for you automatically.
Alongside it, the free Ghost Job Detector scores a pasted listing against 11 weighted signals, with posting age and repost frequency together making up about a quarter of the weighting. Read it as a prompt to look harder rather than a verdict. Levels.fyi is worth a tab for a different reason: its compensation data is free to browse, and a posted salary band that doesn't match what a company has actually been paying is a reason to slow down.
What none of these tools can verify
A posting with an internal candidate already selected is indistinguishable from a real one. It's fresh, specific, complete, and it will pass every signal any detector applies, because the listing itself is genuine. Only the outcome is decided. No tool can see this, and it's a large share of the problem.
The same goes for a req that was real yesterday and was frozen this morning. Detection works on the listing's metadata, and metadata lags reality by weeks.
The word verified also needs pinning down, because the industry uses it loosely. Greenhouse Verified, launched in 2024, recognises employers for their treatment of candidates during hiring. Its CLEAR partnership from 2025 verifies the identity of the candidate. Neither confirms that any particular job exists. A verified badge on a company tells you something real, just not the thing you're checking for.
Three checks worth doing by hand
Search the exact job title plus the company name and look at how many times the posting appears with different dates. Serial reposting with no change to the text is the clearest available signal, and it takes about a minute.
Go to the company's own careers page rather than the aggregator. If the role isn't there, it's stale wherever else you found it. If it is there, note the date it shows, because the aggregator's date is frequently a resyndication date rather than the original.
Check whether the hiring manager exists and has posted about the team recently. A role attached to a person who's been publicly hiring is a different proposition from a listing with no human behind it anywhere.
What this does to your daily number
Filtering cuts your list, and the reasonable response is to send fewer applications rather than to backfill the gap with worse targets. If a fifth of the postings you were going to apply to were never real, then removing them raises your reply rate before you change a word of your resume. Sharpening the resume itself is still worth doing, it just isn't what's costing you here.
The practical version: spend the first ten minutes of a search session culling rather than applying. Kill anything over 30 days old, anything you've seen before, and anything that isn't on the company's own site. Then apply to what's left, properly.
If culling leaves the list too thin, widen by sector rather than relaxing the freshness rule, since the postings you just removed won't get better by being reconsidered. Hiring freezes drive a lot of ghost postings, and the fields that keep hiring through a downturn tend to carry fewer of them for that reason. The applications you don't send are the cheapest improvement available to you.