Nolina
← All resources

TRANSPARENCY REPORT · Q3 2026

Outbound Governance Benchmark, Q3 2026: Baseline and Measurement Method

This first report publishes the measurement contract before a qualifying production cohort exists. Unavailable outcomes are marked not reportable—not converted into reassuring zeroes.

THE SHORT VERSION

Key takeaways

  • Nolina does not yet have a qualifying production cohort for outcome percentages.
  • The metric contract is published now so future reports cannot redefine success after seeing the data.
  • Test fixtures and self-test sends are excluded from market-performance metrics.

A benchmark becomes marketing fiction when the denominator changes after the result is known. Nolina is publishing the metric definitions before publishing performance percentages. As of August 25, 2026, the product has verified system and provider behavior but does not have a qualifying production cohort of buyer outreach.

Q3 reporting status

MetricStatusWhy
Signals processedNot reportableNo defined production cohort
Claims blockedNot reportableFixtures excluded from market metrics
Messages rewrittenNot reportableNo qualifying live-review denominator
Ambiguous sends reconciledNot reportableSelf-tests are not buyer sends
Suppressions caught at dispatchNot reportableTest cases prove behavior, not prevalence

Not reportable does not mean zero. It means the observation does not yet meet the cohort and provenance rules below.

Definitions fixed before measurement

Signal processedA candidate completed source, entity, and freshness evaluation.

Claim blockedA draft-specific assertion failed its evidence right before approval.

Message rewrittenA blocked assertion was replaced with supported or audience-level language.

Ambiguous sendThe provider may have acted but the outcome was not confirmed.

ReconciledProvider evidence resolved an ambiguous send to sent, not sent, or human review.

Suppression caughtA touch otherwise eligible was stopped by the dispatch-time suppression gate.

Cohort and exclusions

  • Include only live buyer opportunities after product activation.
  • Exclude fixtures, demos, self-tests, provider probes, migrations, and recovery drills.
  • Deduplicate by opportunity and touch before aggregation.
  • Publish counts with denominators; never publish a rate without its base.
  • Suppress small slices that could reveal a customer, sender, or recipient.
  • Version any definition change and restate prior periods when comparable.

Threshold for the next edition

The first outcome edition will require at least 500 processed opportunities and 100 live touches across at least three distinct operating lanes. Until that threshold is met, Nolina will publish engineering assurance separately from market-performance data.

This page will remain at the same URL and show a visible last-verified date. Method changes will be appended rather than silently replacing history.

COMMON QUESTIONS

Frequently asked questions

Why publish a benchmark with no production percentages?

Publishing definitions first prevents selective reporting later. The report distinguishes unavailable data from zero and states the cohort threshold required for the next edition.

Are test runs included?

No. Fixtures, recovery drills, contract probes, and self-test messages verify system behavior but do not represent buyer outcomes.

When will Nolina publish counts?

After a qualifying production period contains enough live opportunities and sends to avoid exposing individual activity and to make rates interpretable.

Will the raw data be public?

Only aggregated, non-identifying data will be published. The report will document exclusions, denominators, and any method changes.

CONTINUE READING

Related resources

Editorial MethodologyHow Nolina labels facts, estimates, and unavailable data.AI Outbound SalesSee the operating controls these metrics evaluate.VerificationInspect the product evidence standard.

SOURCE NOTES

Sources and further reading

  1. Artificial Intelligence Risk Management FrameworkNIST

Vendor-authored sources are used to document definitions, workflows, or platform policies—not as independent proof of product performance.

TURN THE FRAMEWORK INTO A SEARCH

What does a real buying moment look like in your market?

Run a pilot