Author: GeoParrot Team
-
The Verdict: A “4.2x More AI Citations” Quality-Score Claim Didn’t Just Overshoot. It Pointed Backwards.
Monday we found a widely-repeated “quality score predicts 4.2x more AI citations” study never built a non-cited control group. We built one. This closes the loop: our own skeptical predictions — which already bet against the marketed claim — still weren’t skeptical enough. 2 of 5 locked predictions confirmed outright (and in a more extreme…
-
Klook SEO & GEO Audit: From Travel Discovery to Korean Support
An evidence-first audit of Klook’s travel marketplace: real category, support and Korean FAQ screens, plus what its sitemap, llms.txt and crawl controls reveal.
-
Notion GEO Audit: Inside Its Scalable Answer System
A visual consulting audit of Notion’s answer system: discovery routes, 475 Help URLs, repeatable page anatomy, structured-data watchpoints, and a practical GEO playbook.
-
One Screenshot Showed AI Mode’s Chat Sidebar on a Plain Google Search Page. It’s the Third Boundary-Blur We’ve Logged This Quarter.
On October 14, 2026, an SEO posted a screenshot of AI Mode’s left chat sidebar appearing on a regular Google web search results page — unreplicated, unconfirmed by Google. Lined up against two other sightings from the same quarter, it’s a pattern, not an isolated test: Google keeps testing ways to fold AI Mode’s interface…
-
We Built the Missing Control Group. Hygiene Score Predicted Who AI Does NOT Cite.
Tuesday we pre-registered a control group the “4.2x citation” claim never had: an 8-pillar hygiene score (G) on every page in the Google AI Mode candidate pool, cited and non-cited. The core bet (a small, positive gap) failed — the gap is negative. Cited pages scored 0.55 average G; non-cited pages scored 0.74. High-hygiene pages…
-
Google Added a New Model to AI Mode Yesterday. The Last Swap Broke Citations for 48 Hours — and No Tracker Is Built to Notice.
On October 13, 2026, Google added Gemini 4 Argon to AI Mode’s model picker for Pro and Ultra subscribers. Six weeks earlier, the last model Google added to that same menu — Gemini 3.8 Flash — shipped with a bug that dropped source links from AI Mode answers for roughly 48 hours. We checked whether…
-
Google Just Opened a Bypass Around the 30-Day Crawl Delay It Quantified Last Week. Most Publishers Can’t Use It.
On October 8, 2026, Google launched the UGC Fresh Data Program — a direct-submission pipeline for approved forums and social platforms, built around exactly the ~30-day crawl-refresh lag Google itself quantified days earlier. We separate what Google confirmed from what we’re inferring about what this means for AI citations.
-
Cited vs. Never Cited: Our Pre-Registered Test of the Control Group a 4.2x Citation Claim Never Had
Monday we found that a “quality score predicts 4.2x more AI citations” study audited only pages that were already cited — no non-cited control group. This week we build the one it’s missing: 10 topics, two engines, and for each query a candidate pool of 10 organic pages scored on an 8-pillar hygiene checklist, labeled…
-
Google Search Console’s Core Report Is Running 44+ Hours Behind. Your “Live” AI Overview Numbers Aren’t Live.
As of October 11, 2026, Google Search Console’s Performance report — the data pipe behind most free AI Overview and AI Mode impression tracking — was running 44 to 64+ hours behind, with no Google acknowledgment. We separate the confirmed delay from the unconfirmed cause, and explain why this compounds a measurement problem GEO practitioners…
-
GEO Vendors Are Citing a “4.2x More AI Citations” Quality-Score Study. Its Own Data Can’t Test That Claim.
A GEO-16 academic framework claims pages scoring 0.70+ on a 16-pillar quality score get cited 4.2x more by AI engines — and GEO vendors are repeating it as proof that quality scoring predicts citation. We opened the paper. Its 1,100 audited URLs were all pulled from existing citations, with no non-cited control group, so “4.2x”…