Sign in

Running an AI Analysis

How to run an AI readiness analysis on your website and what to expect from the results.

This guide walks you through running an AEO/GEO analysis on your website.

Prerequisites

Before running an AI analysis, ensure:

  1. Your project has been crawled — AI analysis uses crawl data to evaluate individual pages. Run at least one crawl first.
  2. Your site is publicly accessible — The analysis checks domain-level files (llms.txt, robots.txt) that must be reachable from the internet.

Starting an Analysis

  1. Navigate to your project in the Evergreen dashboard
  2. Open Insights > AI (AEO, GEO) from the sidebar
  3. Click the Run AI Analysis button
  4. Wait for the analysis to complete — this typically takes 30-60 seconds depending on site size

What Happens During Analysis

Evergreen performs checks in parallel across two levels:

Domain-Level Checks

These examine site-wide configuration:

  • Fetches and validates your llms.txt and llms-full.txt files
  • Parses your robots.txt for AI crawler rules (GPTBot, ClaudeBot, PerplexityBot, etc.)
  • Checks for MCP server availability

Page-Level Checks

These evaluate individual pages from your most recent crawl:

  • Analyzes heading structure and content hierarchy
  • Measures entity density and readability metrics
  • Checks for BLUF patterns and direct answer formatting
  • Validates JSON-LD structured data
  • Evaluates schema properties (sameAs, knowsAbout, dateModified)
  • Detects AI-isms and generic filler phrases

Reading the Results

After the analysis completes, you will see:

AI Readiness Score Card

A composite score from 0 to 100 with your maturity level (Unaware, Reactive, Structured, Optimized, or Leading).

Category Breakdown

Four category cards showing individual scores for:

  • Technical Accessibility
  • Content Engineering
  • Schema & Structured Data
  • Agent Readiness

Each category shows the number of insights found, grouped by priority (high, medium, low).

Insights List

Actionable findings organized by subcategory. Each insight includes:

  • What was found — The specific issue or opportunity
  • Priority level — High, medium, or low
  • Category — Which pillar it belongs to
  • Affected pages — Which pages the insight applies to (for page-level checks)

Running Follow-Up Analyses

After making improvements to your site, run another analysis to track progress. Your AI Readiness Score will update to reflect changes. Compare scores across analyses to demonstrate improvement over time.

Combining with Other Features

AI Insights work best alongside other Evergreen features:

  • Content Audit — Fix foundational SEO issues before optimizing for AI search
  • Analytics Correlation — Identify pages already receiving AI-driven traffic via GA4 and Search Console
  • Shareable Reports — Export AI readiness data alongside traditional SEO scores for client reporting
PreviousAI Insights (AEO/GEO)NextUnderstanding Your AI Readiness Scores