Accountability
How the Reqbeat Hiring-Demand Index will be graded against official vacancy statistics, published before there is anything to grade — and the misses logged beside the hits once there is.
No comparison has been published. No official series is read here yet, so no reference month has an official number beside ours. Until one does, the count of paired months is zero however many editions we have recorded — the two are different numbers and this page states both.
Frozen and dated before any comparison exists, under method v1.8.0. A scorecard whose rules were written after the results were in is a rationalisation; this one can still embarrass us, which is the whole point of publishing it now.
U.S. Bureau of Labor Statistics — Job Openings and Labor Turnover Survey (JOLTS)
For each reference month in which both series have a print, the Reqbeat Hiring-Demand Index's month-over-month percentage change minus the official series' month-over-month percentage change for that month, in percentage points, together with the share of those months in which the two moved in the same direction. An edition's reference month is the calendar month holding the majority of the 28 days its observation was counted over. The comparison is made against the official series' first published print for the month, never a later revision. Movement is compared, never level: this index counts advertised vacancies in a sample of public postings and does not estimate the vacancy stock an official survey measures.
Three inputs, none of them the definition, and the first cannot be hurried. Series length: the comparison needs 12 reference months with a print on both sides, and the Reqbeat side began accruing with the first edition published under the index definition above — the render surface retains a rolling 30-day window, so no earlier month is observable and none is back-cast. Scope: the index counts postings across every market the corpus covers, while each official series covers one country, so the graded side is never the global index — it is a geography-matched sub-series recorded on the same slot, window and deduplication rule, differing from the global observation in the country filter alone. One such series is recorded per benchmark, and each accrues from its own first print exactly as the global index does. Official prints: nothing on this site reads an official series, so no month has an official number beside it. The accrual toward the first comparison is published at /labs/scorecard as it happens, including the months we have not reached.
The whole-month shift — searched over 0 to 3 months — at which the Reqbeat Hiring-Demand Index's month-over-month direction agrees most often with the official series', reported together with the agreement at every shift tested so a shift that wins by one month is visible as winning by one month.
Everything the accuracy metric above is blocked on, and one thing more: a shift of k months consumes k of the paired months, so the longest shift tested needs that many reference months beyond the minimum. Publication timing is not this figure and will not be reported as it: the index prints a reference month before the official series does, which is a fact about two release calendars rather than a measurement of whether the index moves first, and stating the one as the other is the easiest indefensible claim on this site to make.
Both definitions are frozen on jobs.signalsapi.com/method, and no figure publishes under either until that page's two-person method check clears.
One sub-series per official series' scope, counted on the same slot, window, deduplication rule and agency exclusion as the global index — differing from it in the country filter and nothing else. A scoped read that matches nothing records nothing: a zero would be indistinguishable from a country that stopped hiring, and these ledgers are append-only and first-write-wins.
1 print recorded · 0 of 12 months paired
United States, verified against the corpus on 2026-07-22The corpus render surface stores country as a full name, so this is the exact literal an equality filter matches. Verified against the live surface rather than inferred from a sibling table: the pSEO pages, the comparison family and the report cuts read three different columns, and only one of them is what an index observation is counted on.
No official series is read here yet, so no reference month has an official number beside ours. Until one does, the count of paired months is zero however many editions we have recorded — the two are different numbers and this page states both.
Ledger, published open: series-us.json (CC-BY).
Every dated print of the Reqbeat Hiring-Demand Index recorded so far. The ledger is published open: series.json (CC-BY). It is append-only and first-write-wins, so a re-render cannot re-measure a print already in it.
| Edition | Graded as | Observation | Measured | Method |
|---|---|---|---|---|
| July 2026 | June 2026 (7 of 28 days fell in July 2026) | 2,944,467 | 2026-07-22 | v1.8.0 |
Observation = deduplicated postings counted over the stated window. These are observational counts, published as they are made; the derived index point is the gated figure and travels with the index, not with the observations it is computed from.
No comparison has been published. No official series is read here yet, so no reference month has an official number beside ours. Until one does, the count of paired months is zero however many editions we have recorded — the two are different numbers and this page states both. When there is one, it appears here whether we won it or lost it — that commitment is protocol clause 8 above, and it is the reason this table is empty rather than absent.
Full methodology, frozen and versioned: jobs.signalsapi.com/method.