We've been deep in the world of Answer Engine Optimization, running the CiteMET playbook across our best content. The initial results were solid - we saw users engaging with our AI Share Buttons, and we knew we were successfully seeding our content into AI platforms. But with our team's background in AI, we saw this as just the first step.
We understood that for a Large Language Model (LLM), a single, quick interaction is a whisper. A true signal of authority, one that builds lasting memory, comes from a deeper conversation. We saw an opportunity to transform that initial whisper into a meaningful dialogue.
Introduction
You shipped your SaaS with AI. You ask that same AI "what is my product?" and it tells you "I can't access that page."
That was us. givefeedback.dev is a tool I co-founded to turn messy client feedback into clear tasks. We shipped it fast, won a $100k global competition with it, and still couldn't get a single AI model to describe what it did.
The app worked. The website was a ghost to the same models we built it with.
Here's how we took our Client-Side Rendered (CSR) React app, ran it through the C.M.E.T. Method, and got to 100% AI and search visibility in under 24 hours on cite-met.com.
The Before: A Working App With No Website
We ran an audit on givefeedback.dev. The site worked great for humans. To crawlers it was nothing.
Phase One Audit: givefeedback.dev (Before cite-met)
| Metric | Initial Score | Max Points |
|---|---|---|
| Static HTML Content | 0.0 | 5 |
| AEO Directives (llms.txt) | 0.0 | 2 |
| Structured Data (JSON-LD) | 0.0 | 2 |
| SEO Meta Tags | 2.0 | 3 |
| Total C.M.E.T. Score | 4.0 | 12 |
| Visibility Grade | F | - |
The biggest culprit was Static HTML Content. GiveFeedback is a React app, so JavaScript renders everything client-side. To a crawler, the page is <div id="root"></div> and nothing else.
We scored zero on static content, zero on AEO directives, zero on structured data. Total: 4 out of 12. Grade F.
Deploying to cite-met
Goal: fix the blank-page problem and give crawlers enough signal to cite us.
We pushed givefeedback.dev through cite-met.com. Four steps:
1. Connecting GitHub: Pointed cite-met at our GiveFeedback repo.
2. Automated SSG Conversion: cite-met's build engine turned the CSR app into Static Site Generated output. The Static HTML Content problem was gone after the first build.
3. AEO Directive Injection: cite-met generated llms.txt, sitemap.xml, and a robots.txt that explicitly opens the door to AI bots.
4. Structured Data Enhancements: We added FAQPage and SoftwareApplication JSON-LD so models could parse what GiveFeedback is and what it does.
Repo connected to fully deployed site: under 24 hours.
The After: 12/12 and Live Citations
Re-audit ran the same day:
Phase Impact Audit: givefeedback.dev (After cite-met)
| Metric | Score After cite-met | Max Points |
|---|---|---|
| Static HTML Content | 5.0 | 5 |
| AEO Directives (llms.txt) | 2.0 | 2 |
| Structured Data (JSON-LD) | 2.0 | 2 |
| SEO Meta Tags | 3.0 | 3 |
| Total C.M.E.T. Score | 12.0 | 12 |
| Visibility Grade | A+ | - |
12/12. Grade A+.
The score is one thing. The bigger signal: our analytics started logging hits from ChatGPT-User, Googlebot, and ClaudeBot within hours.
How the Models Responded
We re-tested the same prompts across the major models.
ChatGPT: Ask "What is givefeedback.dev?" and it now describes it as a feedback management tool with accurate features, citing our site. The day before, the answer was "I can't access this page."
Perplexity: Started including GiveFeedback in tool comparisons, pulling pricing and feature data straight from the new build. It cites us next to established competitors.
Claude: Reads our docs and gives implementation guidance to users asking how to wire a feedback tool into their workflow.
First-week analytics:
- 300% increase in organic AI-driven traffic
- First citations appeared within 48 hours
- AI crawler visits went from 0 to 15+ daily
- Direct referrals from AI answers started showing up in the acquisition funnel
The clearest signal wasn't in the dashboard. Users started telling us they found us through ChatGPT or read about us in a Perplexity summary. That channel didn't exist for us a week earlier.
What we took away
If you vibecoded a SaaS with AI, odds are that AI can't read it. Four things stuck with us:
1. The blank-page problem is real. CSR apps are invisible to crawlers. React, Vue, Svelte without SSG: same outcome.
2. Models skip, they don't guess. If your site doesn't expose structured data, models won't invent your features. They'll cite your competitor.
3. The fix is fast. F to A+ in under 24 hours, no rebuild. The tooling already exists.
4. We use it ourselves. cite-met exists because we hit this problem on GiveFeedback first. Running our own product through it is how we know the numbers hold up.
givefeedback.dev went from invisible to cited inside a day. Your site can do the same.
Run Your Free Visibility Scan — score in 60 seconds.
Deploy Your Site with cite-met — same path we took.
