How Aivatar Signal audits work — the methodology.

Every Signal audit produces a foundation score, a four-pillar breakdown, and a prioritized fix board in roughly 60 seconds. Here is exactly what runs underneath: deterministic crawl, AI analysis layer, scoring formula, fix board buckets, readiness gating, and SPA-aware detection.

Start freeSee Aivatar Signal

Two layers — deterministic and AI

Layer 1: Signal fetches your homepage and a sample of pages, parses structural signals, compares against a checklist. Fast, repeatable, doesn't depend on external services. Layer 2: Signal calls a structured-output AI model with the crawl signals and asks for judgement on ICP clarity, offer clarity, CTA strength, trust signals, messaging coherence, AI-search readiness.

The four pillars Signal evaluates

  • Technical visibility — 30% — robots, sitemap, canonical, metadata, schema, RSS, alt text, internal links.
  • Content architecture — 25% — page types, H1 quality, content depth, blog freshness, missing archetypes.
  • Trust and conversion — 25% — ICP clarity, offer clarity, CTA quality, trust signals, conversion paths.
  • AI visibility readiness — 20% — machine-readability for ChatGPT/Perplexity/SGE, citation-readiness, schema coverage.

The fix board: Now, Next, Later

Every finding is bucketed by recommended action type. Fix Now is for blockers and high-severity findings that hurt visibility today. Fix Next is for medium-severity findings that matter once the blockers are gone. Fix Later is for nice-to-haves and forward-looking improvements. The fix board is the artefact, not the score.

Readiness gating for Marketing OS

Audits return Foundation Weak, Foundation Fair, or Foundation Ready. Marketing OS unlocks once your project hits Foundation Fair or better — because publishing content on a broken foundation amplifies the problems Signal would have found.