Living report - updated October 19, 2026
The median tech role stays open 77 days
Almost every job dataset tells you what is open. This one tells you what closed, and how long it took. It is built from 109,868 postings that opened and then disappeared from 1,184 companies' own career pages, plus the 53,051 roles live right now.
77
Median days a role stays live
109,868
Closed postings analysed
53,051
Roles live now
1,184
Companies tracked
How long a posting lasts
Half of all closed roles were gone within 77 days. The middle 50% closed between 34 days and 142 days.
- Under a week4.8%
- 1-4 weeks17.5%
- 1-3 months32.1%
- Over 3 months45.6%
Only 4.8% of roles vanished inside a week, so most listings stay reachable well beyond the first few days after they appear.
By seniority
Median days live, tiers with at least 50 closed postings
- Junior66 days (n=8,882)
- Senior76 days (n=32,039)
- Principal84 days (n=1,701)
- Staff85 days (n=3,456)
- Mid86 days (n=4,580)
Junior roles close in 66 days against 76 days for senior ones. Entry-level listings are pulled down faster, which is consistent with them drawing far more applicants per opening.
By work model
Median days live, models with at least 50 closed postings
- Hybrid67 days (n=26,613)
- Onsite77 days (n=6,604)
- Remote80 days (n=32,790)
Remote roles sit open for 80 days versus 77 days onsite. Remote listings reach a far larger candidate pool, so a longer-lived remote posting says more about how many applications have to be filtered than about how hard the role is to fill.
By function
Median days live, functions with at least 50 closed postings
- People & Talent57 days (n=2,812)
- Operations61 days (n=5,240)
- Finance & Legal69 days (n=3,062)
- Design70 days (n=1,858)
- Marketing & Growth70 days (n=6,844)
- Other72 days (n=47,138)
- Data & ML74 days (n=3,805)
- Customer & Support74 days (n=1,670)
- Sales76 days (n=11,298)
- Product78 days (n=2,702)
- Engineering99 days (n=23,439)
People & Talent postings close fastest at 57 days; Engineering take the longest at 99 days. If you are targeting the slower end, a listing that is a few weeks old is still worth an application.
By country
Median days live, countries with at least 50 closed postings
- United Kingdom48 days (n=9,156)
- Brazil63 days (n=1,286)
- France63 days (n=3,159)
- Canada67 days (n=3,172)
- Germany75 days (n=19,935)
- Ireland76 days (n=1,256)
- Spain77 days (n=2,085)
- India77 days (n=4,916)
- Australia78 days (n=1,568)
- United States84 days (n=33,597)
- Netherlands89 days (n=1,403)
- Singapore91 days (n=1,295)
Who publishes a salary
Share of live roles with a structured pay range, countries with at least 200 live roles
| Country | Disclosed | Live | Rate |
|---|---|---|---|
| United States | 1,353 | 19,674 | 6.9% |
| Canada | 52 | 1,292 | 4% |
| Poland | 20 | 591 | 3.4% |
| Italy | 9 | 461 | 2% |
| Spain | 18 | 999 | 1.8% |
| France | 31 | 1,729 | 1.8% |
| Japan | 11 | 733 | 1.5% |
| India | 28 | 1,940 | 1.4% |
| United Kingdom | 42 | 3,652 | 1.2% |
| Australia | 6 | 665 | 0.9% |
| Germany | 62 | 7,479 | 0.8% |
| Brazil | 5 | 615 | 0.8% |
| Netherlands | 5 | 797 | 0.6% |
| Switzerland | 2 | 517 | 0.4% |
| Singapore | 2 | 681 | 0.3% |
Across every live role, 3.7% carry a structured pay range. United States leads at 6.9%, Singapore trails at 0.3%. We only count a machine-readable field, so pay buried in free text is missed and the true rate is higher everywhere. The gap between countries is the reliable signal, not the level.
Seniority mix
24,417 live roles carry a detectable seniority
- Senior14,023 (57.4%)
- Junior4,056 (16.6%)
- Mid2,911 (11.9%)
- Staff2,437 (10%)
- Principal990 (4.1%)
Senior, staff and principal roles are 71.5% of the labelled market, outnumbering junior openings 4.3 to 1. This is an ATS-sourced corpus of funded tech companies, so read it as how that slice of the market advertises, not as the labour market at large.
Work model
28,923 live roles state or imply a work model
- Remote14,399 (49.8%)
- Hybrid10,599 (36.6%)
- Onsite3,925 (13.6%)
Where the hiring is
All 53,051 live roles, bucketed by title
- Other24,270 (45.7%)
- Engineering12,177 (23%)
- Sales5,191 (9.8%)
- Marketing & Growth2,259 (4.3%)
- Operations2,220 (4.2%)
- Data & ML1,697 (3.2%)
- Finance & Legal1,287 (2.4%)
- Product1,282 (2.4%)
- People & Talent1,234 (2.3%)
- Design839 (1.6%)
- Customer & Support595 (1.1%)
The stack employers ask for
Share of 12,177 live engineering roles mentioning each technology
- python44%
- aws24.1%
- kubernetes20.8%
- go19.8%
- java15.5%
- typescript15.1%
- azure14.8%
- gcp13.1%
- react11.6%
- docker11.1%
- javascript10%
- postgres9.5%
- node8.4%
- rust6.4%
- kotlin4%
Where the roles are
Live roles by country
- United States19,674
- Germany7,479
- United Kingdom3,652
- India1,940
- France1,729
- Canada1,292
- Spain999
- Netherlands797
- Japan733
- Singapore681
- Australia665
- Brazil615
Get the roles before they close
New roles matching your filters, every Monday. No spam, unsubscribe any time.
Methodology and limits
Continuous collection since April 2026
Every figure here comes from JobJam's own collector, which reads the public career pages and job feeds of 1,184 companies through their applicant tracking systems (Greenhouse, Ashby, Lever, Personio, SmartRecruiters, Workable, Breezy and BambooHR) plus a handful of remote-job feeds. Every scan reconciles what each company still lists, so when a role disappears from the source it is stamped closed here. That reconciliation is what makes the timing numbers on this page possible at all.
"Days live" is not time to hire. We measure how long a posting stayed on the company's board. A listing can vanish because it was filled, cancelled, expired, or because the company changed ATS. Treat it as the window in which an application was possible, which is the number that actually affects a job seeker.
The clock starts at the posting date where the source gives us one, and at first collection where it does not. Where we fall back to first collection, a role that was already open when we started tracking its company understates its true age, which biases those medians down. Timing covers roles closed in the last 180 days; any breakdown with fewer than 50 closed postings is dropped rather than shown.
Pulling the other way, and worth knowing before you quote the headline: any snapshot of a job board over-samples long-lived postings, because a role that sits open for months is available to be observed closing for far longer than one filled in a fortnight. Our collection began in April 2026, so roles that had already been open a long time before we arrived are over-represented among the closures we can see, and the headline median is high because of it. Expect this number to fall and then settle as the observation window lengthens. The comparisons between groups are far more robust than the absolute level, because the same bias applies to every group.
Work model and seniority are inferred labels, not declared fields, and every rate above is computed over the labelled subset rather than the whole corpus. Salary disclosure counts only a structured, parseable range. Technology mentions use strict word-boundary matching: a naive substring match tags "go" inside "ongoing" and inflates it by an order of magnitude.
Because the corpus is ATS-sourced, it skews toward funded tech companies and scaleups rather than the whole labour market. Figures refresh daily. Free to cite or republish with a link back to this page.
See also: this week in tech hiring, companies hiring right now and the mid-2026 snapshot study.