What does it take to get cited by AI? We deliberately tested this using Genview on Genview itself. The results were simple.
PR alone → cited temporarily
Content alone → doesn't spread
PR + content → higher likelihood of sustained citation
This article is the record of that experiment.
Why We Ran This Experiment
I had long believed that being continuously cited by AI requires two things working together.
External content: Press releases, media coverage, and external mentions. This becomes the trigger for AI to recognize that a brand exists.
Internal content: Website pages, glossaries, columns, and other content assets. This becomes the foundation for AI to understand what a brand is and continue citing it.
We decided to verify this hypothesis using Genview on Genview itself.
How the Experiment Unfolded
Phase 1: April 2026, immediately after beta V1 launch → not cited by AI
Right after launching Genview as beta V1, we were barely cited by AI. There were no external mentions, and the information AI would need to recognize Genview simply didn't exist.
Phase 2: May 11, 2026, PR TIMES release → cited by AI
To verify whether AI citation would continue on external mentions alone, we deliberately kept site updates to a minimum after the PR release. No glossary, no columns — we maintained a landing-page-centric state and observed what happened.
After observing for several weeks to a month, Genview gradually stopped being cited.
ChatGPT (guest mode) during the period when internal content was not maintained. Genview doesn't appear.
Recognition Gained Through PR Disappears Without Content
After the PR release, Genview began being recommended by AI. But maintaining a state where internal content wasn't growing caused that recommendation to fade.
Recognition gained through PR disappears without content.
A press release is the trigger that gets AI to recognize a brand exists. But making that recognition stick — creating a state where AI cites you continuously — requires maintaining your own website and content. External and internal: when either is missing, continuous AI citation becomes difficult.
Beta V2 Launch: Building the Internal Foundation
Acting on the experiment's conclusion, on June 9, 2026 we followed Genview's improvement proposals and produced a large volume of GEO know-how articles, glossary entries, and experiment columns. We also rebuilt the homepage and completed the trust-building setup (E-E-A-T evidence input), re-launching as beta V2. This content build-out is itself the first record of Genview's improvement proposals being put into practice.
Trust-building page after beta V2 launch. E-E-A-T evidence URLs have been entered.
Current GEO Score
Scores as of June 8, 2026 — measured just before the beta V2 relaunch.
Genview dashboard (as of June 8, 2026 — just before beta V2 relaunch)
AI Platform
Score
Mention Rate
ChatGPT
26
37.04%
Gemini
31
51.85%
Claude
32
44.44%
Perplexity
42
52.94%
Grok
29
40.74%
We track three scores in Genview: overall score, competitive comparison score, and citation score. We'll continue following how these scores change and what shifts in AI behavior that produces.
The external foundation is in place. The internal foundation is built. The conditions are set. Now we observe whether AI truly begins citing Genview again.
What We're Tracking Next
If our hypothesis is correct, AI responses should begin to change again over the coming weeks to months. Next time, we record that change.