AEORank Engine
Deterministic scoring engine behind every AEO Content AI audit. Understand how 53 criteria are weighted, benchmarked, and applied in production.
AEORank is no longer distributed as an open-source package
This page is the public methodology reference for the engine behind AEO Content AI. Scoring system | Benchmarks | Contact
Overview
AEORank is the deterministic scoring engine behind every AEO Content AI audit. It evaluates websites across 53 criteria in five fixed-weight pillars: Answer Readiness, Content Structure, Trust & Authority, Technical Foundation, and AI Discovery.
The current model first normalizes criterion weights inside those pillars, dampens overlap between closely related criteria, applies a topic-coherence gate, and then blends the site-level score with a page-fleet score derived from sampled pages. Audit output includes the overall score, score confidence, split headline scores for foundation and content fleet, and page-fleet metrics.
The methodology is documented publicly through AEO Content AI docs and benchmark pages. 11,068 domains scored, 4,313 Y Combinator startups benchmarked across 15 sectors and 28 categories, and the same criteria metadata shown on this page is generated from the production engine package used inside the monorepo.
Current audit outputs
Every current audit can return more than a single score.
- overallScore blends the site score with page-fleet performance.
- overallScoreConfidence reflects heuristic confidence plus sample representativeness.
- headlineScores split the result into
foundationandcontentFleet. - pageFleetMetrics shows page-type mix, weighted page score, coverage, and representativeness.
Public Access
AEORank now runs as a private/internal engine inside AEO Content AI. For most users, the public entry points are the live audit workflow, the scoring docs, and the rankings pages rather than a standalone package install.
Private access
Approved internal consumers can still integrate the private scoped package, but public documentation no longer treats AEORank as a standalone public install.
Product Integrations
AEORank powers the audit experience across AEO Content AI products and internal workflows.
53 criteria
Every criterion has a fixed effective weight determining its contribution to the overall score. Raw criterion weights are adjusted by confidence and overlap controls, then normalized so the five pillars always land at 45%, 25%, 16%, 9%, and 5% of the model before final page-fleet aggregation. See the full scoring system docs for detailed explanations of each criterion.
This docs page stays in sync
The criteria table on this page is generated from the installed AEORank engine package used by AEO Content AI, so criteria-count and effective-weight changes do not require manual rewrites here.
Benchmark Dataset
AEORank powers the benchmark and rankings system. Publicly, that shows up as live rankings and methodology context on this site. Internally, the engine works against benchmark snapshots covering:
Public benchmark views
Use the rankings pages to inspect live benchmark output. Internal package data snapshots are not distributed from this page.
Internal Product Automation
AEORank is now maintained inside the AEO Content AI monorepo and wired into internal CI, worker flows, Studio, and the Chrome extension through shared workspace packages instead of any standalone repo distribution.
Current distribution model
The supported integration path is AEO Content AI itself, or the private scoped package for approved internal consumers.
Methodology Feedback
If you want to challenge a criterion, weight, or benchmark interpretation, start with the scoring documentation and compare the live rankings output. For implementation access or product questions, contact AEO Content directly.