The cost of a certified failure

One estate. One loop. Two intents.

The subject is an enterprise SaaS company with a digital media publishing operation. This page records one software release on that estate, from the first loop to the close. The operator definition lives on the canonical term page. This page carries the cost: recorded work, recorded hours, and converted figures labeled Illustrative.

Canonical term: Certify-then-reject loop.

Rework, inspection, delay — one living tally.

Every figure on this page reads from the same tally. The counts and hours are recorded; the currency is modeled and labeled Illustrative.

1. Tier 01 · Direct rework

   6

   Engineering cycles recorded as 6 pull requests across 5 loops, plus agent compute, the one-time engineering to build the missing assertion, and re-scoping. Illustrative: $4,706.

2. Tier 02 · Accountable inspection

   27

   27 executive hours spent rejecting artifacts the gate had already passed, plus engineering re-verification across five loops. Illustrative: $13,828.

3. Tier 03 · Compounded delay

   3.7 business days

   3.7 business days of blocked release window, at a deliberately conservative 35% blocked allocation, while a green suite kept certifying the wrong property. Illustrative: $4,662.

One incident window, all three tiers: $23,196. Illustrative.

Modeled at a loaded executive rate of $359 per hour (base-band midpoint × 1.3 employer load ÷ 2,080 hours) and a senior engineering rate of $137.50 per hour. Executive hours sit inside the blocked window, so the tiers never double-count them. Calibration points for your own model; every driver adjusts.

Deterministic counts. Probabilistic conversion.

Conversion rows price 27 executive hours and 3.7 blocked business days at the rates above, labeled Illustrative.

Day-one deliverables the estate owns.

The cost curve bends when the property a human judges is named before a token is spent. Manta Graph™ returns that map on day one, and 60% of this defect class is detectable from it. Avoided cost in year one: $139,178. Payback on the Frontier Diagnostic Sprint: 20 days at $7,500 standard environment, 26 days at $10,000 regulated environment. Illustrative.

The same release, governed.

Name the property the accountable executive judges before the first fix: every element inside its designed bounds on every route. Hand the gate that same property, so a wrong result fails the build on the first loop. Keep a person at the control on the release decision, with a captured frame of each rendered route on the Actuality Receipt™. Loops two through five, and the hours they consumed, never start.

Eight verticals where the same loop compounds.

The case is one estate. The pattern is portable. Wherever an automated gate certifies a property a human later rejects, the same three tiers accrue.

Cost the loop. Then decide.

This is the economics page for the same pattern. It answers what a certified failure costs once a green gate records a pass and an accountable human rejects the artifact. Cost figures stay labeled recorded or still open so a briefing that cites this page cites the labeled figure. The operator page is the home of the term.

A modular controller over the stack you already run.

GW Slate™ holds identity, policy, runtime, and receipts so the property a human judges is the property the gate can assert. Human-at-the-control stays on every irreversible step. Actuality Receipts make provenance, authority, logic, and outcome auditable. The Frontier Diagnostic Sprint maps one estate and runs a Micro-Proof demonstration — $7,500 standard environment / $10,000 regulated environment, 100% creditable toward the GW Slate™ build. A running controller is the GW Slate™ Department build from $25,000 or Sovereign Enterprise, and 100% of the Sprint fee credits toward it.