1. Site understanding
Search Lighthouse crawls the homepage, robots.txt, sitemap, sampled pages, metadata, schema, feed links, and page categories so later recommendations have site context.
Methodology
Search Lighthouse separates site facts, search data, deterministic rules, AI interpretation, fix planning, and verification so every recommendation stays tied to evidence.
Search Lighthouse crawls the homepage, robots.txt, sitemap, sampled pages, metadata, schema, feed links, and page categories so later recommendations have site context.
Google Search Console CSV, ZIP, OAuth rows, and Bing Webmaster exports are normalized into query, page, clicks, impressions, CTR, position, country, device, and date evidence.
Rules and trend history identify whether a site is in Needs Search Data, Not Indexed, Google Testing, Early Growth, or Authority Site before prioritizing fixes.
Deterministic checks find indexability, sitemap, canonical, metadata, low-CTR, zero-click, ranking opportunity, and GEO trust issues before AI interpretation is added.
A diagnosis report can become a fix plan with P0/P1/P2 implementation tasks, acceptance criteria, and briefs for Codex, Cursor, Claude Code, or markdown workflows.
Each task gets a verification path: instant crawl, later search data, manual review, or not verifiable. Work is not considered proven until evidence changes.
Paid monitoring stores snapshots, detects meaningful changes, generates alerts, and only creates new reports when action is needed.
Key framework
Search Lighthouse is not a generic audit score. It turns evidence into a work log that builders can ship and verify.
Report
Diagnosis
Fix Plan
Treatment
Task
Action
Verification
Proof
Why Search Lighthouse Stores History
Most tools show the current state of a site. Search Lighthouse stores enough context to explain whether changes actually worked after a fix ships.
Sample output
Monitoring framework
Start with a public URL scan, then add search data, fix plans, and monitoring when you need deeper evidence.