Skip to main content

Labs · public methodology

Growth diagnostics methodology hub

LoudScale publishes free diagnostic tools with evidence-earned scoring models, progressive reveal, and explicit limitations. This is a methodology index — not a calibrated benchmark scoreboard until golden-set results are published.

High-intent free diagnostics

Scoring principles

  • Evidence-earned: dimensions start at 0, not free 40–70 baselines.
  • Separate score, evidence coverage, and confidence on every result.
  • Critical caps (no CTA, noindex, no clear offer) prevent cosmetic As.
  • Single-page HTML overall max 88 — rare A grades need multi-source proof.
  • Findings include observed evidence, example, owner, and validation steps.
  • Scores are LoudScale models — not search-engine or ad-platform truth.
  • Actual AI mentions require prompt tests (not the readiness checker alone).
  • Content Architecture Gap Checker crawls sitemap + internal pages (capped) — not keyword demand.

Model version changelog

  • 2026-07 · Evidence-earned scoring-core; grade A at 90+; coverage caps.
  • wgg-v2 / lpc / hpg / air · Shared single-page HTML ceiling (88); richer findings.
  • seo-v2 · Evidence-earned SEO checker; noindex hard cap.
  • cga-v3-crawl · Sitemap + multi-page architecture crawl (was homepage-nav MVP).
  • benchmarks · 48 frozen HTML fixtures with expected score ranges in CI.

Continue in LoudScale Agent · All free tools