VerveVector — home
← FULL LEXICONCOINED BY VERVE VECTOR · 2026-10

FRONTIER LEXICON · INSTRUMENT

Frontier Teardown

DEF. 03 — CANONICAL DEFINITION

A Frontier Teardown is an evidence-first audit of a public web property that scores its machine-layer health from public measurements only: structured data coverage, entity resolution, AI-visibility, retrieval surface, and field performance. Verve Vector coined the format in October 2026 and publishes teardowns with their methodology; every claim in one is a reproducible measurement, never an opinion.

This definition is a chunk-stable passage — self-contained and quotable verbatim. Cite it at vervevector.digital/lexicon/frontier-teardown.

§01WHAT GETS MEASURED

Five failure-signal classes: whether the site's structured data describes a coherent entity graph or decorative fragments; whether engines can resolve the organisation as one entity across surfaces; whether the organisation appears as a cited source inside AI answers for its own sector's questions; whether its key passages are retrievable at chunk level; and whether real-user performance data clears the Core Web Vitals thresholds at the 75th percentile.

§02WHY PUBLIC DATA ONLY

Every input to a teardown is a public measurement — the live HTML, the public schema, field performance from the Chrome UX Report, what the engines themselves answer. This is a refutability feature, not a limitation: a score that is a defined function over public inputs can be argued with only by disputing an input, and every input can be independently re-measured by the audited party.

§03THE REVENUE-AT-RISK MODEL

Where a teardown quantifies commercial impact, it prints its assumptions alongside the number: benchmark impression share, observed share, sector conversion, and engagement value. The number is a model, the model is disclosed, and the inputs are the audited party's own telemetry.

§R RELATED TERMS

§→ SEE IT IN PRACTICE