This is the case study we wrote about ourselves, because the easiest GEO test bed in the office was the office. Over a six-week sprint we took the Click2.ai marketing site from roughly twenty pages to 83 indexed URLs, every one of them carrying two JSON-LD blocks (a primary type plus FAQPage), every one of them reachable by AI crawlers without a render step.
What we built
Three layers. Eight national pillar pages (GEO, ChatGPT SEO, Perplexity, Claude, Gemini, AI Overviews, Bing/Copilot, llms.txt). Seven national vertical pages. Forty-two city×vertical pages across South Florida plus ten national metro hubs. Every page is static HTML on an edge-deployed host, with IndexNow firing on every push.
What we can honestly report
IndexNow returned HTTP 200 for the full sitemap on the same day we pushed it. Bing's crawler picked up the first batch within 24 hours. The first formal post-program citation measurement is scheduled for July 9, 2026; we'll update this post with the actual numbers — not projections — the day they're in.
The full write-up
Methodology, the four-constant build, the 20-prompt monthly measurement panel, and the Dataset-schema release live on the dedicated page: ai-citation-case-study.html. It's CC-BY 4.0 so anyone running their own GEO program can copy the panel verbatim.