How do you know a job posting closed? 30% of the closures we first recorded were wrong
A job board is judged by what it stops showing. A listing that stays up after the employer filled the role wastes an applicant's evening, and nothing in a public careers API tells you it happened: there is no closed event. A job is there on Monday and missing on Tuesday, and you have to decide what that means. We keep a daily index of open jobs from about 11,800 company career sites on seven…
Determining when a job posting closes can be challenging, as there is no explicit "closed" event in a public careers API. Job postings can remain on the site even after the employer fills the position. To address this issue, a job board monitored 11,800 company career sites across seven systems. From October 30 to November 1, 2026, they recorded 19,811 closures, but 30% of those closures were incorrect. The original rule of comparing job listings from one day to the next yielded inaccurate results due to several factors.
Firstly, a job may not be removed from the system immediately, even if the company has filled the role. The API may experience timeouts, 500 errors, or rate limits, which do not necessarily indicate that the job has closed. Instead, the job remains open until the next API refresh. Secondly, even when the API returns a 404 error or lists no jobs, it does not definitively mean that the job has been closed; the job may still exist on the company's internal board.
Lastly, due to an issue with paging through the job listings, some jobs might appear twice and others might not appear at all. This led to incorrect closures being reported.
To improve the accuracy of job closures, the job board implemented a new method. They compared the total number of jobs reported by the API with the actual distinct jobs present in the job boards. If the number of distinct jobs is fewer than expected, it indicates a partial read, and the job is not closed. They also addressed the issue of job re-posting, where the same job can be listed with a new ID on the same board.
By matching re-posts based on job ID rather than URL, they were able to reduce the number of incorrect closures. By treating a missing row as a question and conducting a sample comparison against the employer's own board, they improved the overall accuracy of job closure tracking.
Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.