How long do job postings stay open? We measured 524,322 of them.

We track 524,322 job postings from 1,077 companies, read from each employer's own careers site. Saved active records can persist during collection gaps.

This counts only postings that carry an employer-supplied date, and excludes listings flagged as evergreen or continuous-hiring — a subset of everything we track. See how we measure. Last updated 8 September 2026 at 20:15 UTC.

The median job posting has been listed for 32 days.

A quarter have been up 13 days or less. A quarter have been up 69 days or more. The oldest still accepting applications was first posted 5,010 days ago.

Of 524,322 postings we count

50.5%

have been listed 30 days or longer

264,594 of 524,322

28.9%

have been listed 60 days or longer

151,518 of 524,322

19.1%

have been listed 90 days or longer

99,913 of 524,322

Each figure is a share of the same 524,322 postings. These describe the postings listed today, grouped by how long they have been up — not the share of postings that survive a given number of days. See the methodology below for why that distinction matters.

If a job is still posted, is it available?

A live listing means the listing is live. It does not tell you whether anyone is reading applications.

We can measure the first thing and not the second. What we can say: 50.5% of currently-listed postings have been up a month or more, and 19.1% for three months or more. Whether that reflects a slow process, a filled role left up, or a role that was never urgent, we do not know and do not claim to.

Some people call the last category “ghost jobs” — that is their term, not a category we measure.

How long have they been listed?

0-7 days15.3%8-30 days34.3%31-60 days22%61-90 days9.5%91-180 days9.2%181-365 days5.2%365+ days4.4%
Distribution of how long currently-listed postings have been open.
Listed forPostingsShare
0-7 days80,43215.3%
8-30 days179,71134.3%
31-60 days115,59322%
61-90 days50,0339.5%
91-180 days48,1609.2%
181-365 days27,2715.2%
365+ days23,1224.4%

Half of all postings are less than a month old. 4.4% — 23,122 postings — have been listed over a year.

By industry

IndustryPostingsCompanies30+ days
Unclassified227,71050747.9%
Retail73,3601458.1%
Healthcare59,2415644.4%
Technology58,08615960.8%
Finance25,19710446.2%
Industrials17,3432252%
Construction & Real Estate13,2564251.8%
Insurance11,9726156.3%
Pharmaceuticals8,0331344.3%
Consumer Goods6,3421641.4%
Materials4,9231051.6%
Utilities3,1751237.9%
Media & Entertainment1,9841949%
Hospitality1,694640.4%
Energy1,166838.4%

Tracked but not rated

A percentage needs at least 1,000 postings and 5 companies. These are tracked, and their counts are published, but we do not put a rate on them:

Defense & Aerospace has 5,201 postings — ample — from 4 employers; that is 4 companies' habits, not an industry's.

Individual historical examples are withheld. Aggregate statistics above retain their original populations and values.

Methodology

Where the dates come from. Each posting's date is the employer's own, published by their applicant tracking system. We never infer or estimate one. Postings without a real date are excluded rather than guessed at.

How often we check. Collection is scheduled daily, but a failed or incomplete read can leave saved active records unchanged. “Still listed” describes recorded status, not proof of a recent successful check.

What is excluded, and why. Evergreen listings and continuous-hiring titles are excluded from every duration figure — a role a company keeps permanently open is not a role that has been open a long time, and mixing them would inflate every number here.

These are composition figures, not survival rates. Every number describes saved active records, grouped by employer posting age. We did not follow a group of postings from appearance to removal. Our own observation history begins 23 July 2026; the ages above run to 5,010 days because they come from employers' posted dates, not from our window. A survival rate would require the second thing, and we will not publish one until we have it.

Coverage and bias. This is not a random sample of the US labor market. It skews toward large employers with public applicant tracking systems — Workday, Greenhouse, Lever, Ashby. Small employers, agencies, and companies whose boards are not machine-readable are absent. Read the figures as describing the tracked set, which is what they measure.

Using these figures

Free to cite with attribution to Zeltru and a link to this page. See our terms.

Download the aggregate figures (CSV) — the summary, buckets and industry table on this page. Not the underlying postings.

This is the longest answer we have. The shorter ones — how long until you hear back, what a removed posting means, what people mean by a ghost job — are on our answers pages, each built from saved collection records.

Related: the job posting leaderboard, longest-open postings, and how we measure.