# Page Audit

The engine classifies the page into one of 8 types (Local Business, YMYL, E-commerce, etc.), then runs evaluation modules - Traditional SEO, RAG, AI Readiness, Link Audit, Schema, E-E-A-T, Subjectivity Filter - all rule-based, no LLMs.

## Page classification (logic gates)

Before evaluation, the engine classifies the page into one of 8 page types. Each type has its own scoring rules - so a local plumber’s site isn’t penalized for missing academic citations, and a YMYL page is held to stricter authority standards.

- Local Business
  - NAP consistency, geo signals, service area
- YMYL (Your Money Your Life)
  - Health, finance, legal - stricter authority requirements
- SaaS / Software
  - Feature clarity, pricing transparency
- E-commerce
  - Product schema, pricing, reviews
- News / Editorial
  - Datelines, authorship, freshness
- Support / FAQ
  - Answer completeness, structured Q&A
- Entity Root
  - Brand or organization hub pages
- Other / General
  - Default rules when no specific type fits

## What the audit evaluates

After classifying the page (e.g. Local Business, YMYL, E-commerce), the engine runs modules in sequence - no LLMs. Order: Traditional SEO, RAG Optimization, AI Readiness, Link Audit, Schema Engine, E-E-A-T, Subjectivity Filter. Before you start with our [AI rank tracker](/content/features/ai-rank-tracker/index.html), use this audit to fix critical crawlability, schema, and trust gaps first.

**Subjectivity Filter** detects vague marketing language ("world-class", "cutting-edge") and applies a penalty based on fluff density.

| Area | What it measures |
| --- | --- |
| Human Trust | Traditional SEO, E-E-A-T, internal links |
| AI Citation | RAG, AI Readiness, link grounding, Schema; gated by crawlability |
| Contextual Autonomy | Per-section topic persistence, pronoun independence, self-contained nuggets |
| AI Crawlability | Penalties for each blocked AI bot (e.g. GPTBot, PerplexityBot) |
| Schema | Universal + conditional + badging + taxonomy; alignment penalty when markup and visible content diverge |
| E-E-A-T | First-hand experience, credentials, accreditation, conflict-of-interest, visual verification |
| Rendering Readiness | Optional: H1/meta/canonical match, content delta, noscript; only when rendering check is on |

## Why Use the Page Audit?

Rankscale delivers a single, high-level view of how any URL supports both human trust and AI citation. Clear status and actionable recommendations.

### Fix Before You Publish

Rankscale audits key URLs so you can catch crawlability, schema alignment, and content issues before they hurt visibility. Prioritized recommendations you can act on.

### Balance Human & AI

Rankscale surfaces whether a page is over-optimized for one side (ai-trap or hidden-gem). Rebalance for the perfect mix with clear status.

### Deterministic & Fast

Rankscale uses Cheerio, JSON-LD parse, and rule-based evaluation - no LLMs in the audit engine. Deterministic, fast results; optional rendering check when you need it.

## FAQ

### 01 What does the Page Audit evaluate?

Rankscale first classifies the page into one of 8 types (e.g. Local Business, YMYL, E-commerce, News/Editorial). Then it runs modules in sequence: Traditional SEO, RAG Optimization, AI Readiness, Link Audit, Schema Engine, E-E-A-T, and Subjectivity Filter. You get Human Trust and AI Citation evaluation, a status (perfect-mix, ai-trap, or hidden-gem), and prioritized recommendations.