Permanent public scorecard
We point the detector at ourselves first.
The direct-mail industry's most common lie is ZIP-smear: a ZIP-level number republished on every route inside it. This page grades our own envelope on exactly that, and it stays up permanently.
The sellable surface
released cohort 2ogs-ncrd-release-2026-09-07 · announced 07 Sept 2026, 23:35 UTC · read from source 12 Sept 2026, 03:00 UTC
192,981
released carrier routes — the only rows we sell against
233,675
Historical, not live: the figure we used to publish. USPS carrier-route roster identity, 2026-02 roster vintage · counts window 2026-08-25. 40,694 routes came out when the data was held to this standard. We publish no share of the roster.
- Withheld — Withheld upstream on the ZIP-set flat multi-route test. Never served as route-level evidence.186 off-surface
- Missing confidence band — Sparse, model-data-thin routes — 46 rural, 37 highway, 30 city — whose modeled figures arrived without a confidence band. A value without a band is a failed read, not a partial one, and we invent no bands.113 off-surface
- Not in return — Genuine USPS-rostered routes (2026-02 roster vintage, 2026-08-25 counts window) absent from the released 2OGS return. Named, not silently dropped and not silently included. A supply-refresh request to 2OGS is on file.2,156 off-surface
- Gate-rejected — Quarantined by 2OGS before release. Neither ingested, inferred nor back-filled by the relay — absent, never zero.20,012 off-surface
Per-state released rows sum to 192,981 exactly.
Roster context — not the sellable surface
source: National Carrier Route Dataset granularity serve · https://owaggydbtywlmgfwvqtz.supabase.co/functions/v1/ncrd-granularity · vintage acs_2019_2023_5yr_route_apportioned · read 12 Sept 2026, 03:00 UTC
These are the upstream roster's own granularity counts across 233,675 rows. They are not the released cohort, and no measured share above is derived from them.
Latest upstream pull
- Status
- recorded
- Reconciliation
- UNRECONCILED
- Vintage
- acs_2019_2023_5yr_route_apportioned
- Window
- 08 Sept 2026, 18:04 UTC
- Received / matched
- 167 / 20
Snapshot history
Roster-context trend from the granularity serve. Not the released cohort.
Audit changelog
Shares below are roster-context self-checks, not released-cohort figures.
12 Sept 2026, 03:00 UTC · nightly · measured 83.7% · smeared 16.3%
Measured share +0.00pt and smear share +0.00pt vs previous audit. Measured 83.7%, smeared 16.3% of 233,675 routes.
11 Sept 2026, 03:00 UTC · nightly · measured 83.7% · smeared 16.3%
Measured share +0.00pt and smear share +0.00pt vs previous audit. Measured 83.7%, smeared 16.3% of 233,675 routes.
10 Sept 2026, 03:00 UTC · nightly · measured 83.7% · smeared 16.3%
Measured share +0.00pt and smear share +0.00pt vs previous audit. Measured 83.7%, smeared 16.3% of 233,675 routes.
09 Sept 2026, 03:00 UTC · nightly · measured 83.7% · smeared 16.3%
Measured share +0.00pt and smear share +0.00pt vs previous audit. Measured 83.7%, smeared 16.3% of 233,675 routes.
08 Sept 2026, 03:00 UTC · nightly · measured 83.7% · smeared 16.3%
Measured share +0.00pt and smear share +0.00pt vs previous audit. Measured 83.7%, smeared 16.3% of 233,675 routes.
07 Sept 2026, 03:00 UTC · nightly · measured 83.7% · smeared 16.3%
Measured share +0.00pt and smear share +0.00pt vs previous audit. Measured 83.7%, smeared 16.3% of 233,675 routes.
06 Sept 2026, 03:00 UTC · nightly · measured 83.7% · smeared 16.3%
Measured share +0.00pt and smear share +0.00pt vs previous audit. Measured 83.7%, smeared 16.3% of 233,675 routes.
05 Sept 2026, 03:00 UTC · nightly · measured 83.7% · smeared 16.3%
Baseline audit: 83.7% of 233,675 routes measured, 16.3% ZIP-smeared. Future runs compare against this.
The standing disclosure
Provenance classes: A USPS deliverable counts = measured. B population/household/age/race structure = modeled (ZIP+4 to decennial block, placement not yet graded). C income, education, psychographics = modeled permanently; ACS is published at block group and no finer, so no route-level value of these is measured from any source.