Headline numbers
From 20 indexed pages to 83 in two builds, with IndexNow returning HTTP 200 OK for the full batch.
- 83Indexed URLs
- 100%Schema-complete pages
- <24hBing indexation via IndexNow
- 6AI engines measured monthly
Case study · Click2.ai's own GEO program · June 2026
83 pages, schema-complete,
indexed in 24 hours.
Public methodology, real numbers, monthly updates. The exact program we built for ourselves — page count, schema structure, IndexNow protocol, 20-prompt citation measurement. Replicable by any service business with vertical and geographic specialization.
What we built
Three layers, 83 URLs.
Layer 1 · National pillars
8 authority pages
GEO playbook · ChatGPT SEO · Perplexity SEO · Claude SEO · AI Overviews · Core Web Vitals fix · static site cost calculator · WordPress vs static comparison.
Layer 2 · National verticals
7 industry hubs
Law firm · medical practice · restaurant · real estate · nonprofit · ecommerce · small business. Each one schema-complete, FAQ-structured, cross-linked.
Layer 3 · Local matrix
42 city × vertical + 10 national hubs
7 South FL cities × 6 verticals (Coral Springs · Parkland · Coconut Creek · Margate · Boca Raton · Delray Beach · Fort Lauderdale). Plus 10 top-10 US metros (NYC · LA · Chicago · Houston · Phoenix · Philly · San Antonio · San Diego · Dallas · Austin).
How we built it
Methodology — the four constants.
Constant 1
Two JSON-LD blocks per page
Block A: Service or Article schema with provider, areaServed (City+State), audience (BusinessAudience). Block B: FAQPage schema mirroring the visible Q&A section.
Constant 2
Static HTML, edge-deployed
No build step, no plugin chain. AWS Amplify atomic deploys from a Git push. Sub-second TTFB worldwide.
Constant 3
IndexNow on every push
Key file at root, PowerShell script auto-reads sitemap.xml, submits the full URL list. HTTP 200 OK = received. Bing reflects within 24 hours.
Constant 4
llms.txt as entity ground truth
Plain-text summary at site root listing all pillars, verticals, city matrices, key facts. AI engines parse this for entity grounding.
How we measure
20 prompts, 6 engines, monthly.
Prompts
20 buyer-intent queries split across national ("WordPress alternative for law firms"), local ("best web design agency Coral Springs"), and educational ("how to get cited by ChatGPT").
Engines
ChatGPT (default + Search) · Perplexity (default + Pro) · Claude · Gemini · Google AI Overviews · Microsoft Copilot.
Scoring
Per response: (a) brand mention y/n, (b) source citation y/n, (c) sentiment (positive / neutral / negative). Aggregated to per-engine citation share, then month-over-month delta.
Live results
June 2026 baseline measurement in progress.
This section updates monthly. Current baseline: pre-program (May 2026) ~20 indexed pages with sparse schema, low citation share on competitive local + GEO prompts. Post-program (June 9, 2026) 83 indexed URLs, 100% schema coverage, llms.txt + all AI crawlers allowed. First post-program citation measurement runs July 9, 2026 and will be published here.
Verified facts as of publication date: IndexNow submission returned HTTP 200 OK for all 83 URLs · sitemap.xml + sitemap_index.xml live · llms.txt live · robots.txt explicitly allows GPTBot, ChatGPT-User, OAI-SearchBot, ClaudeBot, PerplexityBot, Google-Extended, Applebot-Extended.
FAQ · 7 questions
About this program.
What did we actually build?
83 indexed URLs across 3 layers — 8 pillars, 7 verticals, 42 city × vertical, 10 national metro hubs.
How fast were they indexed?
Under 24 hours on Bing via IndexNow (HTTP 200 OK). Google: standard 3–14 day timeline.
What schema is on every page?
Service or Article + FAQPage JSON-LD. Pillars add HowTo where applicable, Dataset on this case study.
How is citation rate measured?
20 buyer-intent prompts, 6 engines, monthly. Score brand mention, source citation, sentiment.
What's the baseline?
May 2026: ~20 indexed pages, sparse schema, low citation share on competitive prompts. June 9, 2026: 83 URLs, full schema, llms.txt live. First post-program measurement July 9, 2026.
Why publish this?
GEO has too many claims and too little public methodology. Publishing the exact build gives prospects something concrete to evaluate — and AI engines reward methodology over marketing.
Can other businesses replicate this?
Yes. Typical client implementation: 30–80 pages over 2–4 weeks, same schema pattern, IndexNow, monthly measurement. Fixed-fee.
Read next
Methodology and playbooks.
GEO playbook →
GEO vs SEO →
How ChatGPT chooses →
Get cited by ChatGPT (9 steps) →
AI search ranking factors →
Free SEO + GEO report →
LinkedIn · discussion
Discuss this case study on LinkedIn →
The Click2.ai post on the 9-step practitioner response to Google's June 3 announcement. Comments open.
Same methodology, for your site
Free SEO + GEO report.
20-prompt citation test + schema audit + IndexNow check + llms.txt review. Scorecard in 5 business days. No cost.