Skip to content

How old is the job you are looking at

Most open roles on a company careers page are more than a week old, and many are past a month. Here is the age of what you are actually seeing.

Jobzyl3 min read

Figures on this page are current as of

"Posted 2 days ago" is the most-read line on a job board and one of the least reliable. Boards relist. Aggregators stamp the date they found a posting rather than the date it went up. A role can sit open for a quarter and look new every time it resurfaces.

So here is the age of a job market, measured the one way that avoids all of that: straight off company career pages, using the date the employer's own system recorded, across 148,263 open roles at 985 companies. Every posting in this set carries a real date. Nothing here is inferred from when we happened to find it.

Half of what is open has been open three weeks

20 days

median age of an open role

Days between the employer's own posting date and today. A median, so half of every careers page is older than this.

Data as of

Three weeks, and because that is the midpoint, half of everything currently advertised is older still.

How long open roles have been open
Show the numbers
How long open roles have been open
CategoryValue
0 to 7 days22.5%
8 to 30 days46.6%
31 to 90 days25.1%
Over 90 days5.9%

Data as of

The bucket most people assume they are searching, the genuinely fresh one, is 22.5% of what is on the board. Everything else has been sitting there at least a week, and 5.9% of it has been sitting there more than a quarter.

Why old is not the same as dead

The obvious conclusion is that the old ones are stale and worth skipping. Resist it, because the data does not support it.

These postings are live. They were on the company's own careers page the last time we pulled it, which is a far stronger signal than an aggregator's copy of a posting. A role open for ninety days is usually open for one of three reasons: it is genuinely hard to fill, it is a rolling requisition the company keeps open on purpose, or the hiring manager went quiet and nobody took the page down.

The first two are fine. The first is arguably an opportunity, because a role nobody has filled in three months has less competition than the one posted this morning.

What age does not tell you is which of the three you are looking at. We deliberately do not publish a "this posting is probably closed" probability, because our own probe data does not support one. Measured across 2,260 liveness checks, the relationship between a posting's age and whether it was still open held neither between boards nor between age buckets. A number that looks like a prediction but is really an artefact of which rows got checked first is worse than no number at all.

Age is a fact about one posting. A staleness probability would be a claim about the future, and we do not have the evidence for one.

What to do with this

Use the date as context rather than as a filter. Sorting by newest puts you in a queue with everyone else, and the 46.6% of postings in the second bucket are the bulk of the market and perfectly live.

Apply to the old ones anyway, and say so when you do. A role open since spring has a hiring manager who is tired of the search, which is a different conversation from the one you have on day two.

Before you write anything off, check the employer's own page. An aggregator's copy of a posting can outlive the original, and one click gets you to the source.

The date sits on every row in product manager roles in Germany and DevOps engineer roles in Canada, and it is the employer's date wherever the board gave us one.

Method, and the one thing this cannot tell you

Distinct postings, collapsed on title, company and location before counting, so a company relisting one role does not count twice.

Age is the employer's own posting date measured against today. That field is populated on every posting in this census, which is precisely why the census is what this post measures rather than the whole index: across the wider set of boards we read, the field is patchier and the dates are less trustworthy.

The limit worth stating plainly: this is the age of what is currently open, not how long a role stays open. Those are different questions, and the second needs a posting followed from the day it appears to the day it comes down. That is a measurement we are set up to make and have not made yet.

Every figure here comes from the jobs Jobzyl has indexed, and carries the date it was pulled. Subscribe by RSS, or run the search yourself.