Your developer is leaving in two months. The question raised in the steering committee is not whether you will recruit. It is how long the position will stay empty, because that duration is what carries a cost.
Nobody in the company has the answer. It is public, free, and under an open licence.
What the data publishes
The public labour statistics agency publishes a quarterly tightness indicator per occupation, comparing the flow of job postings to the flow of job seekers. A high ratio means many postings for few candidates.
The licence is open, commercial reuse included, with attribution of the source. This means you can extract the line that concerns you, put it in a table, and send it to your board without asking anyone for permission.
The approach, in four steps
| Step | What you do | What you get |
|---|---|---|
| 1 | Identify the occupation code in the public classification that matches the position to be replaced | An identifier, not an in-house title |
| 2 | Download the quarterly file and filter on that code | A series of ratios across several quarters |
| 3 | Read the trend, not the latest point | A direction, which matters more than the level |
| 4 | Cross-reference with your own recruitment history for this type of position | A delay in weeks, which belongs to you |
The fourth step is the one that produces the usable figure. The public data frames it, your history quantifies it. A company that took eleven weeks last time will probably take around eleven weeks, adjusted for the trend.
What this changes in the decision
A position empty for eight weeks and a position empty for six months do not call for the same response. The first is covered internally, with overload that is acknowledged and stated. The second is covered by an external service, and that service is budgeted before the departure, not after.
The worst scenario, the most common one, is waiting three months hoping for a quick hire, then bringing in a contractor in an emergency, at emergency rates.
The limits, which are real
The grain of the public classification is broad. It does not know your technical stack and does not care about your job description.
What the ratio measures is a relationship between recorded flows, not a duration. It does not replace your history, it frames it.
The sentence to remember
An empty position has a monthly cost. The only useful question is how many months.
Sources
-
Labour market tightness by occupation, ratio of job vacancy flow to job seeker flow · Dares, via data.gouv.fr, 12 April 2026
Licence: Licence Ouverte 2.0 · Publication trimestrielle, dernière mise à jour du 12 avril 2026. Réutilisation autorisée y compris commercialement, avec mention de la source.