Are 80% of Jobs Never Posted? Is There a Hidden Market?

Career guides, university career centres and LinkedIn posts repeat it: 80 percent of jobs are never advertised, or 85 percent are filled through networking. The figure travels without its source.

The number is a 1966 leftover. In recent hiring data about half of hires came through the posting, and the applications that compete with yours arrive in its first days.

Where the 80 percent came from: a 1966 survey of 27 companies

A 2023 analysis that followed the claim back through the career-advice literature found its roots in a 1960s book by career counsellor Bernard Haldane and in a 1966 Conference Board pilot survey of 27 companies in Rochester, New York. According to that analysis, about a quarter of the hires in that survey had appeared in the newspaper. "Did not appear in the newspaper" became "never advertised", and by the 1970s the finding had hardened into "80 percent of good jobs are hidden".

The 85 percent version is newer and thinner. It comes from a 2016 LinkedIn article summarising an online poll of about 3,000 self-selected respondents, with no published method. The headline says all jobs; the body narrows the claim to "critical jobs".

A third version borrows credibility from government data. A 2013 study of US job openings found that establishments reporting zero vacancies at the end of a month still accounted for 42 percent of the next month's hires. The authors explain much of it as jobs posted and filled inside the same month, which a month-end snapshot cannot see.

Read as "42 percent of jobs are never advertised", the statistic says something its authors did not.

Each figure is old or misread, and each is repeated without its source.

A cousin of the hidden-market claim says that many postings are fake. That one is unmeasured too. In 2025 the US Congressional Research Service looked for a count of postings with no intent to hire and reported that no official statistic exists; the percentages in circulation come from companies that sell hiring services, with methods they do not fully disclose.

In nearly 250,000 recent hires at Ashby's customers, applicants were the largest source, at 43 to 52 percent

Hiring-software companies can count every application and every hire across their customers. Ashby's 2025 analysis of nearly 250,000 hires at its customers, from 2021 to mid-2025, found that people who applied through the posting made up 43 to 52 percent of all hires.

Referrals were about 18 percent, and candidates the company went out and found were about 16 percent.

Referred candidates did far better per head. In the same vendor's 2025 analysis of 38 million applications, referrals were about 1 percent of applicants, and a referred candidate reached an interview 40 percent of the time against 3 percent for an applicant with no referral. Part of that gap is screening before the fact: the employee who refers someone has already judged them.

Referred candidates reached interviews far more often. Across all of those hires, though, the posting is where the largest share began.

The share is lower for executive roles. They were filled by applicants less often, about 35 percent. TeaKo's position is that at director level and above, who you know carries more weight; below that level, the posting is the main door.

On the job boards TeaKo reads in full, a closed posting was online for about 35 days

TeaKo reads employers' own career sites in the US and Canada. On some job systems the scraper reads a window of recent postings, so a posting that leaves the window only looks closed. The numbers below come from the job boards TeaKo reads in full, Greenhouse, Lever and Ashby, where a posting that disappears was taken down.

As of 6 November 2026, about 31,300 postings on those boards had closed in the previous 90 days with a known posting date. The median one was online for about 35 days. Two in ten were gone within 14 days, and four in ten within 30.

Applications arrive early. A 2024 NBER study of 125 million applications made on Dice, a US tech job board, between 2012 and 2017 found that 45 percent went to postings less than 48 hours old.

The pile behind each posting has also grown. Greenhouse's 2026 benchmark, covering 640 million applications at its North American customers, found applications per posting roughly doubled between 2022 and 2025 at those customers, from about 115 to 244.

Put the three together. A posting that closed had typically been online for about five weeks. Almost half of its applications arrived in its first two days, at least on the one board where that was measured. The posting stays up long after the crowd has arrived.

And the pile is larger than it was in 2022, at least at Greenhouse's customers.

TeaKo's view, resting on those three facts: a posting stays up for weeks, but it is read in its first days. The hidden job market is a posting's first days, before the pile behind it has grown.

The myth moves hours away from the postings that close fastest

The 80 percent story does its damage through time.

If most jobs are hidden, the rational plan is to spend evenings on coffee chats and connection requests and treat applying as the long shot. If the posting is the main door, every hour spent chasing an introduction is an hour not spent on a posting that fits, during the days it is being read. That trade-off is the cost of the myth.

The research on networking says which contacts help. A randomised 2022 study of more than 20 million LinkedIn members, published in Science, found that acquaintances were more useful for landing a new job than close contacts, and that the type of connection mattered more than the number.

Networking works, through weak ties. The hours it takes still come from somewhere.

How to use this: fit first, then date

Step 1. Decide fit first: the role, the level, the tasks you can show evidence for. A fast application to the wrong posting is still the wrong posting.

Step 2. Among postings that fit, sort by date and work from the newest. On the one board where it was measured, almost half of applications arrived in a posting's first two days; a posting that is a month old is still open on the boards TeaKo reads in full, but the pile behind it is already there.

Step 3. Apply within the first two days of a posting that fits. As a rule of thumb, a referral is worth asking for when it costs an hour; when it would mean a week of waiting, apply first. Ask the contact how their programme treats someone who has already applied, since some only credit a referral made before the application.

Step 4. Keep networking for the people who are one weak tie away, former colleagues and classmates you know but rarely speak to. Give the rest of your search time to postings.

The posted market is large. The window is short.

As of 6 November 2026, more than 390,000 jobs were open on the employer career sites TeaKo reads in the US and Canada. The posted market is large.

What is scarce is the time each posting stays open. Spend your search inside it.

If fit comes first and date second, the next useful step is a job list built every morning from employers' own career sites, ranked against your resume, with the date each job was posted or first seen. That is what TeaKo's job matching does.