Blog

Should you apply to a job that’s been up for months? What 11,000 open data jobs show

Scroll any job board and you are looking at what is left after everyone else applied. We took the 11,039 US data jobs listed on employers’ own sites on 4 November 2026 and split them by how long each had been up. The old ones are not just older. They are a different set of jobs.

26%
of listed data jobs had been up more than 90 days
25%
of those were remote, against 31% of new ones
29%
sat at employers that had posted nothing new in three months

The remote jobs go first

  • Remote
  • Hybrid
  • On-site
  • Under 30 days
  • 30 to 90 days
  • Over 90 days
Work setting of listed jobs by how long they had been up, among jobs that say where the work is.

Among jobs under a month old, remote, hybrid and on-site are close to a three-way split. Among jobs over three months old, remote falls to 25% and on-site rises to 40%. Remote postings come down fastest (we measured how fast), so the longer a board sits, the more on-site it gets.

Old postings look like they pay more, but mostly don’t

  • Under 30 days$157K
  • 30 to 90 days$161K
  • Over 90 days$188K
Middle stated salary, among listed jobs that state pay.

Across the board, the middle salary on postings over 90 days old is $188K, against $157K for new ones. That gap comes from who leaves jobs up. Compare the same employer’s new and old postings for the same role, and most of it goes away.

Each dot is one employer and role with stated pay on both new and old postings. Old ones paid more at 29, less at 23 and the same at 2. The middle difference was $7K, against $31K across the whole board.

The employers that leave postings up for months are mostly big, well-paying ones. 182 of Meta’s 737 US data postings had been up more than 90 days. For an employer like that, an old posting is often kept open for the next hire.

Some old jobs are at employers that stopped posting

Of every 100 postings over 90 days old, 29 were at employers that had not posted a single new data job in three months: 827 postings at 523 employers.

A five-month-old posting at an employer that adds new data jobs every week is different from the same posting at an employer that has added nothing since summer. The first is often a standing opening. The second is more likely a job nobody is actively filling.

How to read an old posting

Look at the employer, not just the date. Is it still posting new jobs? Does it usually leave postings up for months? Our job pages show both: how fast that employer usually takes its jobs down, and a note when it has posted nothing new in three months.

Check the employer’s own site. A copy on a job board can outlive the original by weeks. Check a job finds the employer’s own posting and when it first appeared.

Don’t give up on remote because the list looks thin. It looks thin because remote jobs go quickly. Look at the newest ones first and check often.

How we counted

These are the 11,039 US postings for data analysts, analytics engineers, data engineers, data scientists, and machine learning and AI engineers that we saw listed on employers’ own hiring systems on 4 November 2026. A posting’s age is counted from the day we first saw it, or from the day it was reposted. For an employer we started reading recently, a posting can be older than we can tell, so ages here are a floor.

Work setting shares leave out the jobs that don’t say where the work is. Pay is the middle of the stated annual range, and only 17% of these postings state one. A same-employer comparison needs stated pay on both a new and an old posting for the same role at the same employer, which 54 pairs had.

An employer counts as having stopped posting when the newest data job we have seen from it is more than 90 days old.