Editorial · Data · Published 2026-04-30

Ghost jobs in remote tech: how many "open roles" are reposts of themselves

Roughly 13.6% of listings in our live remote-tech corpus share their role title and employer with another listing already in the same index. We call those clusters reposts — and they are observable, reproducible, and large.

Last reviewed 2026-04-30 · Data from /repost_clusters.php (refreshed every 10 minutes from the live corpus)

The numbers, right now

8,891

Listings analysed

Live remote-tech corpus

13.6%

Are repost clusters

Share of corpus that is a duplicate of another listing

412

Repeated roles

Unique (title, company) pairs appearing 2+ times

16

Chronic clusters

Same role advertised 5+ times

What we mean by "repost"

A repost cluster is a set of listings sharing both an employer and a normalised role title within the same corpus snapshot. We strip parenthetical scope hints like "(Remote)" and "(EU only)" before grouping, because those vary across boards but describe the same role.

Reposting is not necessarily deceptive. Some legitimate reasons a single role appears multiple times: a company has separate req IDs per region; an ATS export creates duplicate entries; the role was filled and reopened. We report the data; we don't issue verdicts.

What the data does show is how often a remote job-seeker is asked to evaluate the same opportunity twice. For someone applying to twenty roles a week, that's hours of duplicated effort.

Tiers

  • Duplicate — same role appears 2 times. Often legitimate.
  • Persistent — 3 to 4 times. Usually means the listing is being kept fresh on a board.
  • Chronic — 5+ times. The role has been a fixture; either an exceptionally hard hire, an "always-hiring" funnel, or a stale listing nobody owns. We can't tell which from the data alone.

Top clusters right now

RoleCompanyCountTier
Loading clusters…

Cluster table refreshes from the live corpus on every page load. Numbers may shift hour-to-hour as new listings ingest and old ones expire.

Reproduce the analysis

The data behind this page is a single public endpoint:

# Top 20 clusters across the whole corpus
curl -s 'https://remnavi.com/repost_clusters.php?top=20' | jq

# Filter to a single employer
curl -s 'https://remnavi.com/repost_clusters.php?company=Acme&top=10' | jq

# Compare against the live raw corpus
curl -s 'https://remnavi.com/jobs_api.php?limit=1' | jq '.total'

The endpoint is open, uncached for journalist use (10-minute CDN at most), and CORS-enabled. If you want a structured dataset, we publish the underlying corpus daily — see About RemNavi for the API docs.

What this complements

Two related pieces in the same auditor series:

Press contact for context, comment, or methodology questions: press@remnavi.com. We turn around requests in 24 hours when reachable.