About one in five job postings is a ghost job, meaning it is advertised with no intention of hiring anyone from it. Not one in three, which is the figure repeated everywhere and measures something else entirely.
The distinction matters because it changes what you do. One in five is a tax on a job search that you can partly avoid with checks that take two minutes a posting. One in three would be an argument for giving up on job boards altogether.
What counts as a ghost job
Three different things get the same name, and only one of them is dishonest.
The pipeline advert. No role is open, and the company is collecting applications against a vacancy it expects to have later, or simply keeping a pool warm. Applications are read, occasionally, and nobody is hired from them this quarter.
The already-filled posting. The role was filled, internally or by a referral, and the advert stayed up. Some of this is required: a number of employers must post externally for a set period even when the decision is made. Some of it is nobody remembering to take the listing down.
The signalling advert. The posting exists to make the company look like it is growing, to reassure customers or investors, or to remind current staff that they are replaceable. Nobody will be hired because nobody was ever going to be.
From the outside all three look identical, and all three waste the same hour.
One in five postings, not one in three
Greenhouse, an applicant tracking system used by more than 7,500 employers, measures this directly from its own platform: between 18% and 22% of jobs posted through it in a given quarter are ghost jobs. It also found that close to 70% of its client companies had posted at least one in a single quarter.
The widely quoted "one in three job listings is fake" comes from a different study asking a different question. Clarify Capital surveyed employers and found roughly one in three admitted posting a listing with no intention of hiring. That is one in three employers, not one in three postings.
The two numbers are compatible and they are not interchangeable. A company that posts a hundred roles a year and puts up two ghost jobs counts once in the employer figure and twice in the postings figure. Reporting the employer number as though it described listings inflates the problem by about half, and it is the version that has spread.
The Greenhouse figure is the better one for planning because it counts postings, which is the unit you actually apply to. It has its own limits: it covers one platform's client base, which skews towards mid-size and larger companies with formal recruiting, and a ghost job on a platform that never advertised the role at all would not appear anywhere in it.
Why employers post jobs they will not fill
The employers surveyed gave reasons that are mostly not malicious and are all expensive for you.
Half said the company is always open to new people, which is the pipeline advert described as a policy. Just over 40% said the posting made the company look like it was growing. Around a third said they kept it up in case an irresistible candidate applied.
None of those involve deceiving you as a goal. All of them treat your application as free, because to the employer it is: reading it costs a recruiter a few seconds, and not reading it costs nothing at all. The cost sits entirely on your side of the transaction, which is why the practice persists without anybody deciding to be dishonest.
Ghost jobs cluster in some industries and not others
The rate is not spread evenly. In the same platform data, construction postings were ghost jobs 38% of the time, art roles 34%, legal 29% and corporate services around 31%, against the 18% to 22% across everything.
If you are searching in one of the heavier sectors, close to two in five postings may not convert regardless of what you send. That is not a reason to apply less. It is a reason to expect a lower response rate than the published averages and to not read it as a verdict on your application.
Seven checks that take two minutes
How long has it been up? Most boards show a posted date, and a role listed for more than about six weeks with no change is either hard to fill or not being filled. LinkedIn and Indeed both show this; the company's own careers page usually does not.
Has it been reposted? A listing that reappears every 30 days with the same text is being refreshed to stay near the top of search results. Search the exact job title plus the company name and look for older copies of the same advert.
Is there a salary band? Several states require one. A posting from a company that publishes bands on its other roles and not on this one is worth a second look, because a band requires somebody to have agreed a budget.
Is the title evergreen? "Software Engineer, all levels" and "Sales Representative" with no team, product or region named are the shape a pipeline advert takes. A real vacancy usually belongs to a specific team with a specific gap.
Does the posting name anyone? A hiring manager, a team lead, a recruiter with a name. Pipeline adverts rarely do, because no individual owns the outcome.
How many applicants, against how long it has been open? Where a board shows an applicant count, a four-week-old posting with 900 applications and no status change is a queue nobody is working.
What is the company doing otherwise? A hiring freeze, a recent layoff round or a quarter of falling revenue and forty open postings are not consistent. News about the company is the cheapest of these checks and the most informative.
None of these is conclusive on its own. Three of them together is a strong enough signal to spend your effort elsewhere.
What the checks cannot tell you
Every signal above is circumstantial, and the false positives are real. A role can sit open for two months because it is genuinely difficult to fill. A company can post without a salary band because the state it is in does not require one. A small employer may have no recruiter to name.
There is also no way to distinguish a pipeline advert from a role that will open in six weeks, because those are frequently the same posting. An application to a company you would like to work for is not wasted merely because the current listing is a pipeline, provided you are not applying to eight of them.
Treat the checks as a way of ordering your effort rather than as a filter. The question is which twenty applications get the version of your resume that took forty minutes, not which postings are morally legitimate.
What a fifth of postings being unfillable does to your rate
Ghost jobs are already inside every published interview rate, because the people who logged those applications did not know either.
Applications sent with a generic resume convert to an interview about 2.07% of the time. If a fifth of those applications went to postings that could never convert, the rate against real postings is nearer 2.6%. Tailored applications move from 4.23% to about 5.3% on the same adjustment.
Your true rate is therefore better than the headline numbers suggest, which matters when you are judging whether the applications are working at all. The gap between a good search and a bad one is also smaller than the raw rates imply, because part of the failure was never yours to control. The rates themselves and what moves them are in how many applications it takes to get an interview.
The practical response is not to apply less. It is to stop spending tailored applications on postings that fail three checks, and to move effort towards the routes where the posting is not the entry point at all: a referral, a recruiter you have spoken to, or a company careers page reached directly rather than through an aggregator.
What to do when you suspect one
Apply anyway if it is cheap. A generic application to a suspected pipeline advert costs three minutes and occasionally works, because pipelines do eventually convert.
Do not tailor for it. Forty minutes of rewriting spent on a posting that fails the age, repost and salary checks is the single most wasteful thing in a job search.
Do not apply repeatedly to the same company. Your first application to an employer converts at about 6%, and by the eighth it is under 2%. Blanketing a company whose postings look like pipelines is the worst version of that.
Log it either way. You cannot tell which of your applications went to ghost jobs, but you can tell which companies never respond to anything. After three, treat that employer's board as noise.
Track your own rate rather than the published one. Thirty applications is enough to know whether the problem is the market, the postings or the application. The runway that gives you time to find out is covered in how long you can afford to job hunt.
Frequently asked questions
What is a ghost job?
A job posting advertised with no intention of hiring from it. It covers three different situations: a pipeline advert collecting applications against a future vacancy, a role already filled whose listing stayed up, and a posting that exists to make a company look like it is growing.
What percentage of job postings are fake?
Between 18% and 22% of postings, measured directly from one large applicant tracking platform. The "one in three" figure quoted widely is the share of employers who admit posting one, which is a different denominator and roughly half again as alarming.
How can you tell if a job posting is real?
Check how long it has been listed, whether it has been reposted, whether it carries a salary band, whether the title names a specific team, whether anyone is named as the contact, the applicant count against the age of the posting, and whether the company's recent news is consistent with hiring. Three failures together is enough to redirect your effort.
Why do companies post ghost jobs?
Surveyed employers said they keep a pipeline open as policy, that postings make the company look like it is growing, and that they stay up in case an outstanding candidate applies. Reading an application costs the employer almost nothing, so the practice continues without anyone deciding to deceive.
Should I still apply to a suspected ghost job?
Send a generic application if it costs you three minutes. Do not spend a tailored one, and do not send repeated applications to the same employer, because your first application to a company converts roughly three times better than your ninth.
Which industries have the most ghost jobs?
In the same platform data, construction was highest at 38% of postings, followed by art at 34%, corporate services around 31% and legal at 29%, against 18% to 22% overall.