Ghost Jobs in 2026: How to Tell a Real Requisition From a Pipeline Trap

The Numbers, and Why They Disagree So Wildly

Search this topic and you will be told, in the same afternoon, that 18%, 27%, and 47% of listings are fake. Those figures are not competing measurements of one thing; they measure different things with different instruments. Knowing which is which is the difference between calibrated caution and paranoia.

The honest summary is a range, not a headline: somewhere between one in five and one in four listings is unlikely to produce a hire in the near term, with higher estimates depending on counting duplicates and pipeline postings as fakes. That is bad enough without rounding it up. Greenhouse president Jon Stross, whose company's own data sits at the low end of that range, described the state of the market to the Wall Street Journal in blunter terms than the percentage suggests.

The Requisition Behind the Posting

Every posting you see is the public face of a requisition record inside an applicant tracking system, and requisitions come in types. The most important type to understand is the evergreen requisition: a req deliberately left open with no end date, used to accumulate a pool of pre-screened people for high-volume or hard-to-fill roles.

Here is the detail that reframes the whole debate. In Workday's own documentation for recruiting administrators, an evergreen requisition is used to gather a candidate pool but cannot be used to make an offer or process a hire. Candidates have to be moved onto a regular requisition first. Greenhouse publishes best-practice guidance for the same pattern so that always-open roles report consistently. So when people say a posting is "not a real job," a large slice of the time they are describing a req that is genuinely real, genuinely staffed by recruiters, and structurally incapable of hiring anyone until it is converted.

That matters for what you do next. An application into a pipeline req is not wasted effort — it is deferred effort. It writes the same durable, searchable record every other application writes, which is exactly the mechanism described in how recruiters search the ATS database for past applicants. What it will not do is give you an interview this month. Price it accordingly: ten minutes, not ninety.

Signals That Actually Discriminate

Most "spot a ghost job" checklists list symptoms that real jobs share. A vague description is weak evidence; plenty of real postings are written badly. These are the signals with some diagnostic power, and the ones that only look like signals.

The single cheapest check is the careers page. Aggregators scrape, cache, and keep serving listings long after the source is gone. If the posting is not present on the employer's own ATS-hosted board — the URL will usually carry a vendor fingerprint such as boards.greenhouse.io, jobs.lever.co, myworkdayjobs.com, or an iCIMS subdomain — then what you are looking at is most likely a stale copy rather than a conspiracy. Apply through the source or not at all.

The Verification Layer Employers Started Shipping

2026 is the first year job seekers get machine-readable evidence instead of vibes. Greenhouse ships Greenhouse Verified, a set of badges displayed on the job post itself. There are four categories — Fair, Prepared, Respectful, and Communicative — earned by actually using features that correspond to each behaviour: inclusion reminders, interview-specific questions, job-specific attributes, anonymisation, candidate surveys, email stage notifications, and the share of applicants still active. Verifications refresh daily, and respectful and communicative are available on all product tiers while fair and prepared require the Expert tier.

Read that list carefully, because it tells you what a badge does and does not prove. Communicative is tied to stage notifications and to how many applicants are still moving through the funnel — which is real, checkable evidence that a pipeline is alive. None of the four is a promise that this specific requisition will convert into a hire. A verified badge raises your prior; it does not settle the question.

LinkedIn has taken the parallel route of verification on the hiring side, tagging listings where the poster or organisation has been verified, and reports that more than half of its listings now carry some verification. Same caveat: it verifies who posted, not whether a seat exists. The same distinction between identity and intent shows up on the employer's side of the market too — see AI auto-apply bots vs the ATS, where fraud detection and identity verification are being pointed at candidates for exactly the mirror-image reason.

The Law Is Starting to Catch Up

The most concrete change in 2026 is regulatory. As of 1 January 2026, Ontario's publicly advertised job posting rules took effect under the Employment Standards Act for employers with 25 or more employees. Two provisions matter directly here:

• Vacancy disclosure. A publicly advertised posting must state whether it is for an existing vacancy. That is a legal requirement to label the pipeline posting as a pipeline posting.
• The 45-day rule. If an employer interviews an applicant, it must inform that person whether a hiring decision was made within 45 days of the last interview. Employers must also retain postings, application forms, and the information given to interviewed applicants for three years.

New York has a bill on disclosure of expected fill dates, and New Jersey and Pennsylvania have pending bills on hiring timelines and transparency. None of that is settled law yet, and none of it applies to the bulk of US postings today. But the Ontario text is the template everyone else is now drafting against, and it targets precisely the ambiguity this article is about: not whether a company is lying, but whether it told you which kind of posting you were looking at.

A Routine That Costs You Ten Minutes

1. Find the source posting. Search the company's own careers page for the title. Not there? Treat the aggregator listing as stale and move on.
2. Read for the pipeline tell. "Future opportunities," "talent community," "we are always looking for," an all-levels title with no team. In Ontario postings, look for the explicit existing-vacancy statement.
3. Check the age and the repost pattern. A posting live and unchanged for three months with a fresh applicant count every week is a pool, not a seat.
4. Look for verification. Greenhouse Verified badges on the post, or LinkedIn's verification tag. Present raises your prior; absent proves nothing.
5. Tier your effort, do not binary it. Confident it is real: full tailoring pass. Ambiguous or pipeline: ten-minute version of your master resume, correct keywords, accurate answers. Obvious fiction: skip.
6. Answer the questions yourself, always. A ghost job you got filtered out of on a wrong answer is still a wrong answer on your permanent record — see ATS knockout questions.

The reaction that wastes the most time: treating a suspected ghost job as a personal insult, then either ninety-minute-tailoring it out of spite or abandoning the whole search for a week. Both convert a 20% base rate into a 0% one.
The reaction that compounds: ten minutes, accurate answers, correct keywords, submitted, logged, forgotten. Your record enters a database that 46% of sourced hires now come out of, and your remaining eighty minutes go to the two postings you verified are real.

Frequently Asked Questions

Are 47% of job listings really fake?

That figure is at the top of a wide range and depends on counting duplicates, aggregator copies, and pipeline postings as fakes. Greenhouse's own platform data puts it at 18-22%, and listing-level analyses of LinkedIn have landed near 27%. A defensible working assumption is one in four or five — enough to change how you budget your time, not enough to justify quitting the job boards.

Is applying to a pipeline posting a waste of time?

No, but it is a different transaction. You are buying a place in a searchable candidate database rather than an interview this month. That has real value — pipeline reqs exist precisely because recruiters pull from them later — provided you spend ten minutes on it rather than ninety.

Does a Greenhouse Verified badge mean the job is real?

It means the employer is measurably doing specific things: notifying candidates at each stage, keeping applicants moving, structuring interviews. That correlates with a live process. It is not a guarantee that this requisition converts to a hire, and the fair and prepared badges are only available on higher product tiers, so their absence is not a red flag.

Why would a company post a job it cannot fill?

The documented reasons are mundane rather than sinister: building a candidate pool, satisfying a policy that roles be advertised, benchmarking salaries, reassuring staff or investors that the company is growing, and simple failure to take down a filled req. Malice is rarer than administrative inertia.

Should I just apply to everything to compensate?

That is the trap. Ghost jobs lower your response rate, and spraying lowers it again — applications per hire have already tripled since 2021, and employers have shipped fraud detection aimed at mass submission. The arithmetic and the counter-measures are covered in AI auto-apply bots vs the ATS.

What actually raises my odds on the real postings?

Making sure the file you send parses into the profile you intend, and that it ranks on the criteria the posting names. If you are new to how that pipeline works, start with what an ATS is and how it works, then the ranking mechanics in how ATS match scores are calculated. Keeping your LinkedIn profile aligned with your resume also matters more in a ghost-job market, because inbound recruiter search is the channel that bypasses postings entirely.