How Long a Posting Stays Open Depends on the Platform, Not the Employer
Based on 445,429+ analyzed job listings · Updated 2026-09-04
We started this piece expecting to name employers who run ghost jobs. We are not publishing that piece, because the number we would have reported turned out to be measuring something else. Once we controlled for how often each source actually gets re-checked, the pattern that survived was not about which employers post fake jobs. It was about which platforms enforce closure at all. Postings first seen 65 or more days ago, 64,250 of them across ten major sources, close within 60 days 99.6% to 100% of the time on direct-ATS platforms like Ashby, Lever, and SmartRecruiters. On Workday, 69.6% are still marked active past day 60. Both numbers are checked often enough to trust. The difference is the platform, not the employer using it.
See where you actually stand in this market
Upload your resume and Seeker compares you against current openings — which roles fit your experience, and which are screening you out.
99.8%
Closed within 60 days on Ashby
Re-verified on 28 of the last 30 days, so this number is checked often enough to trust
100.0%
Closed within 60 days on SmartRecruiters
Re-verified at least once every day of the last 30
69.6%
Still marked active past 60 days on Workday
Checked on 21 of the last 30 days, a cadence close enough to Himalayas that the gap between them looks real
3 of 10
Sources we chose not to trust yet
Checked too rarely in the last 30 days to tell real longevity from a stale crawl
Share of postings closed within 60 days, by platform
Bars scaled to 100%
Limited to sources re-checked on most of the last 30 days. Three sources checked on 5 of the last 30 days are excluded here; see the full table and methodology below for why.
All ten sources we measured, including the ones we discounted
| Platform | Closed by day 60 | Re-checked in the last 30 days | Read this as |
|---|---|---|---|
| SmartRecruiters | 100.0% | Every day, often more than once | Real: the platform closes fast |
| Ashby | 99.8% | 28 of 30 days | Real: the platform closes fast |
| Lever | 99.6% | 30 of 30 days | Real: the platform closes fast |
| Greenhouse | 85.2% | Every day, often more than once | Real: still fast, less absolute |
| Workday | 30.4% | 21 of 30 days | Real: checked nearly as often as Greenhouse, so this much lower rate looks genuine |
| Himalayas | 19.8% | 21 of 30 days | Real: same cadence as Workday, a genuine platform difference |
| Reed | Unclear, 82.2% stuck "stale" | 5 of 30 days | Discounted: checked too rarely to separate closure from crawl lag |
| Oracle Cloud | 0%, all still "active" | 5 of 30 days | Discounted: checked too rarely to separate closure from crawl lag |
| Apple careers | 99.6% still "active" | 20 of 30 days | Held back: cadence looks adequate but the result did not survive a first check, needs more before we'd publish it |
"Closed by day 60" is the share of postings first seen 65 or more days ago (n=64,250 across all ten sources) no longer marked active in our copy at the 60-day mark. Re-check frequency counts distinct days in the last 30 with at least one re-verification for that source. A source checked on only 5 of the last 30 days cannot tell you whether a posting is genuinely long-lived or simply has not been looked at recently, so we did not use those two rows as evidence either way.
Source: Seeker job corpus, 445,429 active job postings, snapshot 2026-09-04. seekerscore.com/insights/ghost-jobs-are-a-platform-effect-2026
The direct-ATS platforms close fast, and we can trust that
Ashby, Lever, and SmartRecruiters are checked nearly every day, and 99.6% to 100% of postings on them are gone within 60 days. That is not a small sample or a lucky window: it holds across 64,250 postings first seen at least 65 days ago. When a source is checked this often, a low active-rate at day 60 is not an artifact of our crawler missing something. It means the platform, or the employers who use it, actually closes reqs.
Workday and Himalayas are checked at the same rate and still disagree
This is the comparison that makes the finding a finding rather than a coincidence of crawl schedules. Workday and Himalayas are both re-checked on 21 of the last 30 days, the same cadence. If the gap between them were a crawling artifact, checking them equally often should erase it. It does not: 69.6% of Workday postings are still active past 60 days, against 19.8% on Himalayas. At matched cadence, that is a real difference between how the two platforms and their typical postings behave, not a measurement error.
Two sources we threw out, and one we are holding back
Reed and Oracle Cloud were checked on only 5 of the last 30 days. Reed shows 82.2% of postings stuck in a "stale" state and Oracle Cloud shows 100% still "active": neither number tells you anything about the posting, only about how long it has been since we last looked. We are reporting both here so the exclusion is visible, not silent. Apple's careers page is a harder case: it is checked on 20 of the last 30 days, a cadence that should be adequate, yet 99.6% of its postings still show active past 60 days, higher than anywhere else we trust. We do not yet know if that reflects Apple's actual hiring behavior or a quirk in how that source's status field updates, and we are not willing to publish a number we cannot explain. It is left out of the chart and the headline claims above.
What this means if you are reading a stale-looking listing
A posting that has been up for months is not, by itself, evidence that a company is not really hiring. It is much stronger evidence about which platform it lives on. A 90-day-old Ashby or Lever posting is unusual enough to be worth a second look. A 90-day-old Workday or Himalayas posting is closer to normal for that platform and tells you comparatively little on its own.
How we measured it (and what we did not claim)
We took every posting first seen 65 or more days ago across the ten sources with the most volume, 64,250 postings, and checked whether it was still marked active in our copy 60 days after first-seen. To rule out the obvious confound, being marked active only because nobody had looked recently, we separately counted how many of the last 30 days each source had at least one re-verification recorded, and required a source to be checked on most of those days before we treated its active-rate as a real signal. Two sources failed that bar and are shown but excluded from the headline numbers. A third, Apple, passed the cadence bar but produced a result we could not reconcile on a first check, so we are holding it back rather than publish it caveated. We are not claiming this list ranks every ATS or job board, only the ten we had enough volume to check, and we are not claiming a platform's closure rate reflects deliberate policy versus incidental default settings; we only measured what happens, not why the platform is built that way.
Frequently asked questions
Does a job posting staying open for months mean it is fake?
Not by itself. Which platform hosts the posting predicts how long it stays open far more than anything about the employer. A 90-day-old posting on Ashby or Lever is unusual for that platform; the same age on Workday or Himalayas is closer to typical.
Which job platforms close postings fastest?
Among the ten sources we checked with enough re-verification frequency to trust the result, Ashby, Lever, and SmartRecruiters close 99.6% to 100% of postings within 60 days. Workday (69.6% still active past 60 days) and Himalayas (80.2%) are checked at the same rate as each other and still disagree, which is what makes the difference between them a real finding rather than a crawl artifact.
Why didn't you name which employers post the most ghost jobs?
Because the pattern we found is a platform-level effect, not an employer-level one, and two of the ten sources we looked at were checked too rarely for us to trust any number we could report about them. We would rather publish a smaller, defensible finding than a bigger one we cannot stand behind.
See where you stand
Upload your resume to see your skill gaps, match scores, and which roles fit your background.
Analyze my resumeRelated Data
We Analyzed 1,835 “Entry-Level” Job Postings. Most Still Demand Experience.
“Entry-level” / junior postings analyzed: 1,835
The Highest-Paid Skills in 2026, From 73,374 Job Postings That Disclosed Pay
Job postings we analyzed: 73,374
Backend Engineer Market: What 200,000 Job Listings Tell Us
Active backend roles: 24,600+
The 25 Fastest-Growing Skills in 2026 Job Postings
Jobs analyzed: 133,000+
Guides
Dataset & citation
- Dataset
- 445,429+ active job postings
- Last verified
- 2026-09-04
- Methodology
- How we measure
Suggested citation
SeekerScore. "Ghost Jobs Are a Platform Effect, Not an Employer One." Analysis of 445,429+ job postings, 2026-09-04. https://www.seekerscore.com/insights/ghost-jobs-are-a-platform-effect-2026
Data derived from Seeker's job corpus of 445,429+ listings across multiple sources. Updated 2026-09-04. Individual results vary based on resume content, target market, and role specifics.