{
  "registry_version": "1.0.0",
  "data": {
    "title": "Auto-Close Stale Job Requisitions",
    "status": "active",
    "evidence_ids": [
      "evidence.job_seekers_beware_of_ghost_jobs_clarify_capital_survey_of_hiring_managers"
    ],
    "content": "# Auto-Close Stale Job Requisitions\n\nImplement automated ATS policy to expire public listings when headcount authorisation lapses or no interview activity occurs within 60 days.\n\n### Expected Effects\n- mech.stale_or_orphaned_job_requisition stops operating: a listing cannot outlive the authorisation that funded it\n- Silence after applying acquires a bound — the record is closed and the closure is sent\n\n### Measurements\n- `stale_requisition_rate`\n- `closure_latency_days`",
    "aliases": [
      "I-001"
    ],
    "id": "int.auto_close_stale_job_requisitions",
    "type": "intervention",
    "summary": "Implement automated ATS policy to expire public listings when headcount authorisation lapses or no interview activity occurs within 60 days.",
    "targets": [
      "mech.stale_or_orphaned_job_requisition",
      "bar.application_ingestion",
      "pat.closed_then_reposted_requisition_motif",
      "loop.inflated_requirements_search_saturation_loop"
    ],
    "actor": "employer-policy",
    "scope": "organizational",
    "cost": "low",
    "evidence_level": "supported",
    "expected_effects": [
      "mech.stale_or_orphaned_job_requisition stops operating: a listing cannot outlive the authorisation that funded it",
      "Silence after applying acquires a bound — the record is closed and the closure is sent"
    ],
    "measurements": [
      "stale_requisition_rate",
      "closure_latency_days"
    ],
    "specimens": [
      {
        "kind": "ats",
        "label": "The rule, and what it did",
        "lines": [
          {
            "at": "config",
            "text": "Policy: expire_listing when headcount_authorisation lapses OR no interview activity for 45 days",
            "tell": false
          },
          {
            "at": "02:00",
            "text": "req #4471 — last interview activity 47 days ago → listing withdrawn from careers page and aggregators",
            "tell": true
          },
          {
            "at": "02:01",
            "text": "Candidates with applications in flight: 84 — each notified that the search has closed, with the date.",
            "tell": false
          }
        ],
        "reading": "The same batch job that produces silent bulk rejections can be pointed at the listing instead of the candidate. The cost is a configuration line."
      }
    ],
    "perspectives": [
      {
        "actor": "actor.employer_policy",
        "sees": "A configuration line, and the count of open requisitions it would withdraw this quarter, on the same cycle where headcount is reviewed.",
        "reads": "What is being spent is optionality: a listing that closes on a timer cannot be held open while funding is decided, and reopening it runs the approval chain again.",
        "does": "Sets the interval and what counts as activity, and decides whether a lapsed authorisation withdraws the listing or only flags it for review."
      },
      {
        "actor": "actor.recruiter",
        "sees": "Listings dropping off the careers page overnight, and the batch notification going to every application still in flight against them.",
        "reads": "A requisition closed on a timer is counted as closed rather than filled, and the pipeline built against it is gone if the search restarts.",
        "does": "Sends the closure notice with the date on it, and asks for the requisition to be reopened where the manager still intends to hire."
      },
      {
        "actor": "actor.candidate",
        "sees": "A message naming the requisition, the date the search closed, and the reason the listing came down.",
        "reads": "The process ended for a reason that has nothing to do with the application; there is nothing further to wait for and nothing left to infer.",
        "does": "Closes the file, stops holding capacity for a possible next round, and reapplies if the role is posted again."
      },
      {
        "actor": "actor.ats_vendor",
        "sees": "A customer enabling an expiry rule pointed at the listing, on the same scheduler that already expires unreviewed applications.",
        "reads": "This is an existing batch capability with a different target rather than new machinery; whether it runs is a customer setting.",
        "does": "Ships the rule as a configurable option, and its default value decides how many customers ever run it."
      }
    ]
  }
}