The ghost job count you see is a floor, not a rate
As of 17 November 2026, our 30-day window shows 106 ghost-flagged listings out of 19,878 live postings across 15 boards.
That low number suggests remote job boards are mostly clean, but it is misleading because the window excludes older listings, and staleness is the strongest ghost signal. In fact, 6,883 of the 6,989 total ghost-flagged listings in our historical data are older than the 30-day window. The in-window figure understates the historical problem. If you are scrolling through current boards, you are seeing a filtered view that hides the bulk of empty or fake jobs, because the most obvious stale posts have already aged out of the metric you are comparing against.
Staleness hides the real problem #
You might assume that a low ghost count in the last 30 days means fewer people are wasting evenings on dead roles. Our data says the opposite. The 30-day window is a clean cut, but remote job boards are rarely curated that tightly. When we look at the 100,869 listings we have ever ingested, 6,989 are flagged as ghosts. Most of these, specifically 6,883, are older than the last 30 days. This matters because the 106 in-window ghosts are the tip of the iceberg. The low number is not evidence that ghost jobs are rare; it is an artifact of how the window is defined. Staleness is the strongest ghost signal, and the 30-day window excludes exactly the listings where that signal is strongest. So the in-window count is a floor, not a rate. It tells you how many ghosts are currently visible in that specific slice of time, not how many exist in the broader pipeline of remote listings. Each of the 6,883 older ghosts is a listing someone could have wasted an evening on, applying to a role that was already closed, empty, or fake.
What to do before you apply #
Since the boards do not always remove stale posts quickly, you have to check the posting date yourself. If a listing has been live for more than 30 days, treat it with extra caution. We do not have per-board curation data to tell you which boards clean up faster, so assume you will encounter lingering posts on any board. For example, imagine a "Senior Backend Engineer" role that has been live for two months. That is a hypothetical illustration, not a specific listing from our data, but it shows why age matters. The 6,883 older ghosts in our dataset are the warning sign. They represent the historical weight of stale listings that the in-window figure does not capture. If you are in a location that is less served by major boards, you are more likely to encounter these uncurated posts, but the data does not break down ghost rates by geography. What it does show is that the current 30-day view is incomplete. Stop relying on the 106 figure to feel safe. Prioritize newer listings, and if a post is older than the window, verify it with the company before applying. You can see how we track these boards and the method behind the counts at https://remoty.work/job-boards/. We run a remote job board built partly from the sources we measure, so we are not a neutral party, which is why the method is published.
Written by the remoty.work team with AI assistance, from our own board data. Numbers carry the date they were counted. Spot an error? Tell us.