Most jobs are posted on a Tuesday, but only just, and not every week. The data is 117,153 jobs posted on 12,228 company job boards from 19 Oct 2026 to 15 Nov 2026 and still open on 19 Nov 2026. Allowing for the jobs that had closed by then, 20.7% of postings fell on a Tuesday, against 18.6% to 19.9% on each other weekday and 2.9% across Saturday and Sunday. Tuesday led in 3 of the 4 weeks (2 of 4 for US jobs alone). Jobs go up in working hours: 68.2% of US jobs with a posting time went up between 9:00 and 17:00 US Eastern time.
Download the numbers: CSV ยท JSON
Jobs posted on each weekday
The index was built at 04:03 UTC on 19 Nov 2026 from 23,287 company job boards holding 461,985 open jobs. In the last four full weeks before that, Monday 19 Oct 2026 to Sunday 15 Nov 2026, 12,228 of those boards posted the 117,153 jobs counted here, so every weekday appears four times.
| Day | Jobs | Raw share | Adjusted | Each board once | US, Eastern |
|---|---|---|---|---|---|
| Mon | 20,962 | 17.9% | 18.9% | 18.4% | 20.0% |
| Tue | 23,523 | 20.1% | 20.7% | 21.1% | 20.8% |
| Wed | 21,521 | 18.4% | 18.6% | 19.1% | 18.3% |
| Thu | 23,830 | 20.3% | 19.9% | 19.4% | 19.2% |
| Fri | 23,542 | 20.1% | 19.1% | 19.6% | 18.5% |
| Sat | 1,964 | 1.7% | 1.5% | 1.3% | 1.3% |
| Sun | 1,811 | 1.5% | 1.4% | 1.0% | 1.9% |
All columns count jobs still open on 19 Nov 2026. Raw share is in UTC. Adjusted corrects for jobs that closed before the snapshot (below). Each board once gives every company board one vote, so a board that posted hundreds of jobs counts as much as one that posted a single job. US, Eastern is the adjusted share of the 61,239 US-located jobs in US Eastern time. For the 25,641 Europe-located jobs in Central European time (CET), Tuesday also leads (20.9%), ahead of Thursday (20.6%).
Why the raw count is not the answer
A snapshot only holds jobs that are still open, and a job posted on a Monday has had four more days to be filled or taken down than one posted that Friday. That makes later weekdays look busier. In the raw count, Thursday leads with 20.3%, a hair ahead of Tuesday and Friday at 20.1%.
The data shows how large the effect is. The number of jobs from the same weekday falls 13.9% for each week of age, about 2.12% a day. After weighting each day's jobs by that rate, Tuesday leads with 20.7%. Tuesday also leads when each board counts once (21.1%), for US jobs in Eastern time (20.8%) and for European jobs in CET (20.9%). Across all 456,597 open jobs, whatever their age, Tuesday and Wednesday tie at 20.8%. Monday to Wednesday together take 58.1% of the adjusted count.
Week by week the lead is less steady. Tuesday led in 3 of the 4 weeks; in the week from 9 Nov 2026, Friday did, when Lever boards posted far more than usual on Thursday and Friday. For US jobs and for European jobs, Tuesday led in 2 of the 4 weeks each.
So the familiar advice that Tuesday is the top posting day holds, but the margin is under one percentage point and one busy week can change it. For a job seeker, the useful pattern is broader: new jobs arrive on every weekday at a similar rate and almost never at the weekend.
Posting day by ATS
| ATS | Jobs | Top day | Weeks led | Weekend | By board |
|---|---|---|---|---|---|
| Greenhouse | 52,601 | Tue 21.5% | 3 of 4 | 1.9% | Tue |
| Lever | 14,796 | Fri 21.1% | 2 of 4 | 5.4% | Fri |
| Ashby | 17,439 | Thu 21.2% | 2 of 4 | 3.4% | Tue |
| Recruitee | 9,387 | Mon 21.3% | 1 of 4 | 1.1% | Tue |
| Personio | 4,625 | Tue 20.9% | 2 of 4 | 1.4% | Tue |
| Workable | 18,305 | Thu 20.2% | 3 of 4 | 3.9% | Tue |
Each ATS's own closure rate is used. By jobs, Tuesday tops only Greenhouse and Personio, and Greenhouse holds 52,601 of the 117,153 jobs, so it weighs most in the pooled result; counting each board once, Tuesday tops five of the six ATSs. Weeks led counts the four weeks in which the top day also led on its own; within one ATS the top day changes often, so read close shares as a tie. By board names the top day when every board counts equally. Lever's weekend share comes mostly from boards that post many jobs: with each board counted once it is 2.2%, against 5.4% by jobs.
What time are jobs posted?
US jobs go up through the US working day. Few appear before 8:00 Eastern time (0.6% in the hour from 7:00), then 6.4% arrive between 8:00 and 9:00. From 10:00 to 17:00 every hour holds 7.8% to 9.4%, and the busiest three hours run from 12:00 to 15:00 (27.9%). The often-quoted "post at 11 AM" is not a peak here. European jobs follow office hours in CET: 71.6% between 9:00 and 17:00, with the most at 15:00 (10.3%). Workable gives only a date, so it is left out of the hours.
For a job seeker, that makes one daily check late in the working day the efficient habit: by 17:00 Eastern time, 78.1% of a day's US postings are already up, and by 17:00 CET, 78.4% of a day's European ones.
For job-feed builders: when to refresh
If a feed or job board re-reads career pages once a day, the refresh time decides how long a new job waits to be picked up. Across the five ATSs with posting times, postings peak at 15:00 UTC (8.5% in that hour). A daily refresh at 23:00 UTC gives the shortest mean wait, 8.6 hours, and picks up 38.0% of jobs within 6 hours. At 13:00 UTC, just before the US rush, the mean wait is 15.3 hours. A refresh at 19:00 UTC picks up the most within 6 hours, 45.9%. Our own index is built at about 04:00 UTC: a mean wait of 11.4 hours.
Mean wait from posting to the refresh
Method and limits
- Source: the ATS Jobs Search index built at 04:03 UTC on 19 Nov 2026 from employers' public job feeds on Greenhouse, Lever, Ashby, Recruitee, Personio and Workable. It covers the boards that crawl found and could read, not every employer: employers on other systems, such as Workday or iCIMS, are not in it, so the shares describe these boards rather than the whole job market. The board and job totals are its size, not ATS market share.
- Posting time: the vendor's own field: Greenhouse
first_published, LevercreatedAt, AshbypublishedAt, Recruiteepublished_at, PersoniocreatedAt, Workablepublished_on(a date only). Lever's and Personio'screatedAtis when the job was created, which can come before it was published. - Copies and undated jobs: a job with the same title, location and posting time on two boards of one ATS counts once. That removed 5,304 jobs, 4,757 of them on Recruitee. With the 84 jobs that had no date also left out, 456,597 of the 461,985 open jobs remain.
- Closure adjustment: the daily rate is fitted from how each weekday's count falls from week to week, then each day's jobs are weighted by it. If posting volume itself changed over the four weeks, the rate absorbs that too.
- Region: read from the location text. Of all 456,597 open jobs, 46.5% name a US place and 23.1% a European one; the other 30.4% name another region, several regions or no place we recognize. Of the 117,153 jobs in the four weeks, 61,239 are US-located and 25,641 Europe-located. US Eastern and CET include their daylight-saving changes on 1 Nov 2026 and 25 Oct 2026.
- Reproducible: build_data.py rebuilds the CSV and JSON from the index files, refuses any other index build by its SHA-256 hashes, and fails if a number on this page is not in that data. The index files are not published; the CSV and JSON carry every count used. This page was drafted with AI tools from that data.
Search these jobs yourself
ATS Jobs Search searches the same index by title, location and posting date; with its "only jobs not returned before" option, a scheduled run becomes a daily feed of new jobs. Developers can read the feeds directly with the open-source ats-job-feed library. Scaleturn makes both; these are not affiliate links. How reliable those public feeds are at scale is on our ATS job posting API crawl data page, and which AI job titles these employers post is counted from the same index, as is how old open job postings are.
Sources
ATS Jobs Search job index, built 19 Nov 2026, summarized in job-posting-days-2026-11.csv and job-posting-days-2026-11.json.
Posting-time fields as read by the index from each vendor's public job feed.