Discovery: publication and crawl access
The first four checks cover unauthenticated 200 responses, robots.txt, sitemaps and WAF or bot controls. Verify access with real user agents and server logs, not policy text alone.
- Priority pages return 200 without authentication
- robots.txt and meta robots do not conflict
- Sitemaps contain canonical URLs only
- CDN and WAF allow verified crawlers
Understanding and extraction: page-level clarity
The next eight checks cover entity names, legal operator, expertise, page topic, concise answers, headings, internal links and visible-schema consistency. Each URL should stand on its own.
- Keep company, brand and location naming consistent
- Give each URL one primary topic
- Place a direct answer near the beginning
- Connect related services and articles semantically
Trust and measurement: evidence to iteration
The final eight checks cover original evidence, official sources, update dates, accountability, corrections, cited pages, grounding queries and referrals. Citation count is not rank or authority, so evaluate it alongside qualified pipeline.
- Verify claims with original or official sources
- Show publication, update and editorial responsibility
- Log cited pages and associated topics monthly
- Connect AI referrals to qualified inquiries