Unlisted, a service that monitors job postings on employers’ own career sites, reported in September 2026 that 163,057 dated postings had been open for more than 90 days. That represented 28.3% of the dated postings in its catalog. The median open posting was 36 days old, while 94,106 postings had been open for more than 180 days.
The report’s official title graphic identifies the source of the September 2026 stale-posting figures.
The report calls postings older than 90 days “stale,” but deliberately does not label any individual listing a “ghost job.” That distinction matters: an applicant tracking system is software through which an employer publishes and manages vacancies, and a posting can remain live while a company collects candidates for several openings, fills similar roles continuously, or keeps an evergreen position available.

The workplace photo fits the report’s explanation that an older listing may remain open while employers collect candidates or fill similar roles.
The age patterns vary substantially. Hospitality had the largest share over 90 days, at 43.9%, while healthcare had the smallest, at 19.7%. Listings associated with Lever showed a 48.2% stale share, compared with 17.2% for Workday, although the report does not establish that the software itself causes those differences. Among countries with the most open postings, the reported share ranged from 21.5% in Australia to 40.9% in Sweden.
Removal data adds another useful comparison. Of 88,895 postings removed from employer sites during the preceding 30 days, 13,013, or 14.6%, came down within a week; 30.8% closed within two weeks. Separately, 3,915 of 94,333 closed postings—4.2%—were republished within 30 days for the same role and employer.
For job seekers, the central limitation is that posting age alone cannot reveal whether interviews are continuing. Unlisted says it reads public applicant-tracking-system APIs every few hours and counts listings still open on employer sites, but it cannot determine hiring intent or track how long the same vacancies remain on LinkedIn or Indeed. Its closure figures are also constrained by the report’s observation period. The data therefore documents freshness on employer sites, not a definitive count of real or fake vacancies.