On 6 of 35 measured sources no posting carries a date at all, and on others the reported date lands on the day we crawled it. Both break the same number.
The age of a job ad looks like a fact. It is a subtraction: today, minus the date the ad was published. That only works if the date is there and means what it says. In our crawl it often does neither.
We measure publication-date availability once a week, across every posting we hold. In the September 20, 2026 snapshot that is 7,343,843 postings: 92.83% carry a date, 7.17% carry none.
The missing ones are not spread evenly. Of the 35 sources with at least 500 postings in that snapshot, 6 publish no date on any record we hold: successfactors, icims, pinpoint, recruiterbox, comeet, jobvite. Workday, the largest single source we measure, is missing a date on 46.45% of its postings.
An undated posting still exists, and still shows up in search. It just cannot enter any age or staleness measure. So every figure about "ads older than 60 days" is computed on a subset, and how big that subset is depends on which software the employer posts with.
The second failure is harder to see, and we only found it by crossing the source's own date with the date our crawler last saw the posting.
On some sources the two land on the same day, or within a day, far more often than a spread of publishing dates would suggest. Workday: 66.82%. USAJOBS: 62.11%. And on some sources the reported date falls after the day we last saw the ad. Workday again: 27.2%. Paylocity: 25.92%.
We want the claim here to stay smaller than it looks. A posting we crawled on the day it went up produces the same pattern, and we cannot tell those apart. What we can say is what we cannot do: for those postings we cannot separate a publication date the source reports from a date we effectively observed. We do not get to call it the publishing date.
Combine several platforms into one dataset and use posting age as a variable, and the variable is not measured the same way in every row. Some rows carry a date the source states. Some carry a date that may be our own observation. Some carry nothing.
That does not make the number unusable. It makes it a number about postings that carry a date, on the sources that supply one. That is a narrower claim, and it is the one we make.
We claim that the availability of a source-reported publication date varies a lot by source, and that we can measure how much.
We do not claim that a missing date means an old or fake posting, that any platform is hiding a date, or that the field is the day a vacancy actually opened. It is a property of the plumbing, not of the employer and not of the job.
We publish the per-source table, all 35 of them, on the publication-date page.
→ Publication dates · Hiring platforms · Ghost jobs and listing age · How we measure
Want a cut we have not published? A region, an industry, a role family, a time series, one country against another — we can pull it from the same corpus these posts are measured on.
Tell us what you are looking at: [email protected]. We answer either way.
Source & licence. All figures here come from our own crawl of public job postings, measured against the corpus published in the monthly report and its JSON (CC BY 4.0).
Cite as: DeUnemployment, “Two ways a publication date fails you”, https://deunemployment.com/blog/two-ways-a-publication-date-fails-you (data as of 2026-09-20).