A ghost job posting is a job advertisement the employer has no real intent to fill — a role listed publicly while nobody is actually being interviewed, hired, or planned for. On the Greenhouse platform in Q2 2025, 17% of postings met exactly that definition: jobs the business never intended to actually hire for (per Greenhouse data provided to CNBC, 25 August 2025).
TL;DR:
- 17% of jobs posted on Greenhouse in Q2 2025 were ghost jobs; the platform’s broader range is 18–22% in any given quarter (per Greenhouse 2024 State of Job Hunting Report and CNBC, Aug 2025)
- Four in five companies on Greenhouse posted at least one ghost job in Q2 2025; for 9.6% of companies, more than half of their listings were ghosts (per Recruiter.co.uk citing Greenhouse, Sept 2025)
- Independent academic work (Ng, arXiv preprint, 2024, using LLM-BERT on Glassdoor) landed at up to 21% — the same neighbourhood from a different data source
- Junior software developer roles posted in Germany fell 54% between January 2020 and February 2025 (per Indeed data reported by The European Correspondent, April 2026) — fewer real openings, same posting-shaped noise
- No single signal is proof, but the pattern is consistent: 30+ days live, repeated reposts, vague scope, “always hiring” language, and an obviously mismatched title for the company size
What a Ghost Job Actually Is (and What It Isn’t)
The clearest working definition, from Hunter Ng’s 2024 preprint out of Baruch College (CUNY), is a posting “if it does not have an immediate or reasonably foreseeable immediate vacancy” (per Ng, arXiv 2024). The US Congressional Research Service, in its April 2025 report on the topic, uses essentially the same framing: postings “for positions that do not exist, or that employers are not planning to fill immediately” (per Handwerker & Pepper, CRS Report IF12977, 25 April 2025).
That wording matters because it excludes the honest edge case that gets used to hand-wave the practice away: a genuinely hard-to-fill specialised role that the employer would love to fill in six or nine months, kept live to keep the pipeline warm. That’s not what candidates are complaining about. What they’re complaining about is postings where no interview loop exists, no hiring manager has budget, and no decision is coming — listings whose entire purpose is to exist, not to hire.
A peer-reviewed empirical framework for identifying such postings was published in Business Economics (Vol 60, pp. 216–229, 1 Nov 2025), using LinkedIn data via CoreSignal to formalise what had until then been mostly folk wisdom and vendor surveys. The phenomenon has a measurable footprint on labour-market statistics — enough to help explain why the US Beveridge curve, which historically tracked vacancies against unemployment cleanly, has drifted since 2010.
A ghost job is also not a scam job. Scam listings — fake companies harvesting personal data or asking for payment — are their own problem, well-documented by the FTC. Ghost jobs are usually posted by real companies with real HR teams, and that’s the part that stings; you can’t spot them by checking whether the employer is legitimate.
How Big the Problem Actually Is
The most trustworthy numbers come from platforms that see their own back end. Greenhouse reports that “in any given quarter, 18–22% of the jobs posted on the Greenhouse platform are classified as ghost jobs” (per Greenhouse 2024 State of Job Hunting Report, 10 December 2024). Q2 2025’s snapshot came in at 17% — the same order of magnitude (per Greenhouse via CNBC, 25 August 2025). The company-level view is more uncomfortable: four in five companies on the platform posted at least one ghost job in Q2 2025, and 9.6% of companies had ghost rates above 50% (per Recruiter.co.uk citing Greenhouse, September 2025).
Ng’s 2024 Glassdoor analysis, using a BERT-based classifier, put the rate at “up to 21%” — arriving at the same neighbourhood from a completely different data source. The Congressional Research Service is more cautious: its 2025 report states plainly that “there are no official statistics on the magnitude of ghost jobs” and notes that the firms behind most industry surveys “market hiring-related services” and disclose limited methodology, which makes it difficult to assess the quality of published estimates (per Handwerker & Pepper, CRS IF12977, 2025).
The mechanism showing up in the aggregate labour-market data is starker. Revelio Labs’ chief economist Lisa Simon told Connecticut Public Radio that in 2018, roughly 1 in 5 job postings failed to result in a hire; in 2024, roughly 1 in 2 do not (cited in Ng, arXiv 2024). Some of that gap is longer average time-to-hire and some is postings that were never going anywhere. Both are bad news for the person doing the applying.
For an EU-workable remote developer, the underlying market context matters as much as the ghost-job rate on top of it. US tech postings stood 36% below their February 2020 level in mid-2025, with software engineer postings specifically down 49% (per Indeed Hiring Lab, 30 July 2025). Junior developer postings in Germany dropped 54% from January 2020 to February 2025 while senior postings dropped only 15% (per Indeed data reported by The European Correspondent, April 2026). Fewer real openings, the same overall posting noise, and a rising share of that noise never intended to hire in the first place — that’s the reader’s actual environment.
Why Employers Keep Posting Them
The Congressional Research Service report lays out the deliberate motives most cleanly: employers post jobs they have no immediate intention to fill in order to evaluate future candidate pools, signal growth (to existing staff who they want to know they’re replaceable, or to prospective investors), show they interviewed a diverse pool while intending to hire internally, or hunt for extraordinary candidates who might justify creating a role that doesn’t yet exist (per CRS IF12977, 2025). Ng’s 2024 preprint adds a fourth from the empirical data — HR “productivity theatre,” where recruiter workloads look busier when the requisition queue is full and posting costs essentially nothing on the margin because the ATS is already paid for.
Two mechanisms are worth pulling out because they show up as distinct signals on a posting. One is pipeline maintenance for hard-to-fill roles — the honest edge case, which slides into ghost territory when the pipeline is maintained for roles with no headcount attached. The tell is the same senior-specialised requisition up every quarter with no interview activity between reposts. The other is market intelligence: collecting applications reveals live data on salary expectations and applicant skills, useful for benchmarking existing employees’ compensation. This one is the most cynical, because it explicitly extracts unpaid labour — a customised application — from candidates in exchange for nothing.
The CRS also notes that some ghost postings aren’t deliberate at all: third-party job boards routinely copy listings from other sites, and when the employer takes down the original the copy keeps circulating. Staffing agencies sometimes invent openings to demonstrate to employers that they can recruit talented people (per CRS IF12977, 2025).
The reason none of this changes on its own is that the incentives don’t cross the fence. The employer’s cost of a ghost posting is essentially zero; the candidate’s cost is measured in hours of customised writing.
Eight Signals a Listing Is Probably a Ghost
No single item on this list is proof — plenty of real jobs are posted for 30 days, plenty of real jobs are vague, and a genuine startup will sometimes hire above its apparent weight class. What matters is the pattern. Three or more of these on the same posting and the odds shift sharply.
| Signal | Why it correlates with ghost postings |
|---|---|
| Live for 30+ days | About 1 in 7 postings stays active longer than 30 days (per Clarify Capital via Forbes, 9 April 2026); the longer a role sits open in a market this soft, the less likely it’s actively being worked |
| Reposted repeatedly | The same requisition disappearing and reappearing with minor edits is the single most common qualitative complaint from candidates; if a job description reads identically to one you already applied to, it usually is |
| No salary or salary range | Only about 2% of live tech ads located in EU member states state a salary; the DE/AT/CH/UK/IE/NL/EU-remote combined sample is 6.7% (per Jobdex, H1 2026 report). Missing salary alone means little in Europe — but combined with the other signals, it’s another data point |
| Vague responsibilities | ”You’ll wear many hats” without a specific team, product, or codebase named is characteristic of postings written by HR without a real hiring manager attached |
| ”Always hiring” language | Explicit “we’re always accepting applications for this role” is a pipeline-maintenance tell, not the promising employer signal it’s marketed as |
| Recent public layoffs at the company | A company that laid off 15% of engineering three months ago is unlikely to be hiring the same seniority back at the same title; check the news before applying |
| Title/company-size mismatch | ”Principal Distributed Systems Architect” at a 12-person Series A that hasn’t raised in 18 months is almost never a real requisition |
| No response 3+ weeks after application, posting still live | Not diagnostic on its own — plenty of real hiring is that slow — but if the posting is still live after your no-response, the odds tilt further |
What to Do Differently as an EU-Workable Remote Developer
The change is not “apply to fewer jobs.” It’s “spend the customised-writing time on postings that show evidence of being real.” Three moves matter.
First, filter on posting age and recurrence patterns before you draft anything. A listing you’ve seen twice in three months, or one that’s been up for six weeks with no obvious activity, isn’t worth a two-hour cover letter. An aggregator sees this pattern that no single job-board view of one listing can — the same requisition appears across sources over time, and only the vantage point of aggregation makes that visible. If you’re actively hunting remote-EU work, that’s the specific problem an aggregator like Xeito solves that a single ATS view cannot.
Second, apply channels around the front door. The candidates finding remote-EU work fastest in 2026 are the ones combining public applications with warm introductions — an ex-colleague at the company, a direct message to the engineering manager on LinkedIn, or an Ashby/Greenhouse careers page you found by cross-referencing recent GitHub commits or conference talks. A backchannel application is much harder to lose in a ghost-posting review queue because someone inside is watching for it.
Third, treat the initial screen as your own filter, not just theirs. Ask the recruiter three specific questions on the first call: when the requisition was opened, how many candidates are in the pipeline, and when the hiring manager expects to make an offer. A recruiter running a real requisition will answer all three concretely; a recruiter running a ghost posting will hedge on at least two.
Ghost postings are a systemic problem individual candidates cannot fix. What you can fix is the amount of your own time absorbed by them.
Frequently Asked Questions
How can I tell for certain that a specific posting is a ghost job? You often can’t — from the outside, ghost and real postings look identical, which is what makes the practice work. What you can do is score a listing across the eight signals above. Three or more concurrent signals, especially “posted 30+ days” plus “reposted” plus “no salary in the EU,” is enough to deprioritise it and move on rather than customise a full application.
Are ghost jobs illegal anywhere? Not at the US federal level or in the EU as a category. The Truth in Job Advertising and Accountability Act (TJAAA) was drafted by a US advocacy working group in May 2025 as a proposed federal ban but has not been introduced in Congress (per CNBC, 25 August 2025). At the state level, New York’s S8874 passed on 2 June 2026 and would require employers with 100+ employees to disclose whether a listing is for a current vacancy and the expected fill date; Pennsylvania’s HB2321 goes further, requiring disclosure of hire dates and AI use in hiring (per HR Dive, 11 June 2026). The EU Pay Transparency Directive forces salary disclosure but not intent-to-hire disclosure, so nothing in EU law currently targets ghost postings directly.
Is a job that stays up for six weeks always a ghost? No. Genuinely specialised senior roles routinely take 60+ days to fill in tight sub-fields (staff-plus infra, security research, ML systems), and hiring processes at large enterprises can take months. The 30-day flag is a probability shift, not a verdict. Combine it with the other signals before you deprioritise.
Does “no salary in the job ad” mean it’s a ghost in Europe? No — most EU tech ads still don’t publish salary. Jobdex measured the EU-member-state-only rate at around 2% in H1 2026 (with the broader six-country sample at 6.7%). That will change as more member states transpose the EU Pay Transparency Directive, but for now missing-salary is a weak signal in Europe on its own; it matters in combination with the others.
What’s the difference between a ghost job and getting ghosted after an interview? A ghost job is a posting that was never a real hire in the first place. Being ghosted after an interview is what happens when a real hiring process breaks down or the role is de-prioritised mid-loop — 61% of job seekers reported experiencing it in the Greenhouse 2024 report, and 63.5% of German applicants received no reply at all to any of their applications in the previous 12 months in the Indeed/Appinio March 2026 representative survey (n=1,000). Both are failures, but they show up at different points in the funnel and you defend against them differently.
Sources
- Tech worker was frustrated with ghost jobs — now he’s trying to pass a national ban — CNBC (25 Aug 2025)
- “Ghost” Job Postings — Congressional Research Service Report IF12977, Handwerker & Pepper (25 April 2025)
- Greenhouse 2024 State of Job Hunting Report — Greenhouse (10 Dec 2024)
- Ghost jobs, real costs: labor market distortions from ghost job postings — Business Economics, Vol 60, pp. 216-229 (Nov 2025)
- Why is it so hard to find a job now? Enter Ghost Jobs — Hunter Ng, arXiv preprint 2410.21771 (Baruch College, CUNY, Oct 2024)
- The US Tech Hiring Freeze Continues — Indeed Hiring Lab (30 July 2025)
- Finding a job shouldn’t be this hard — The European Correspondent (April 2026)
- The State of European Tech Job Ads, H1 2026 — Jobdex (2 July 2026)
- 1 in 7 Job Postings Are Ghost Jobs, New Study Reveals — Forbes / Rachel Wells (9 April 2026)
- Disturbing number as ghost jobs rise in the UK — Recruiter.co.uk citing Greenhouse (Sept 2025)
- New York passes a bill aimed at halting ‘ghost jobs’ — HR Dive (11 June 2026)
- The Pharma Bloodbath Part V — Ghosting Reloaded — MoreThanCareer citing Indeed/Appinio survey published 23 April 2026 (May 2026)