Pinned Tweet
Website health & AI search visibility in one scan.
GazeRank audits SEO, performance, accessibility, and visibility across ChatGPT, Claude, Perplexity & Gemini—giving you prioritized fixes, not endless metric dumps.
Try a free scan: gazerank.com
Apple is paying Google $1 billion a year to license Gemini for an upgraded Siri launching this spring, per eMarketer. Even Apple chose renting AI over building it.
If a trillion-dollar company won’t out-build this, your business won’t either. Show up inside it instead.
Check out my latest article: How Small Agencies Compare AI Visibility Tools linkedin.com/pulse/how-small… via @LinkedIn
Ahrefs tracked Copilot’s citations across real US queries: Amazon alone holds 14.6% of citation share. Amazon plus Walmart make up nearly a quarter of everything it cites.
If you sell anything, Copilot behaves like a commerce engine wearing a chat interface.
Check out my latest article: What to Check Before Installing SEO Score Tools linkedin.com/pulse/what-chec… via @LinkedIn
GEO blogs claim author bios lift AI citations 40%, 47%, whatever — each number from a different vendor’s study.
Google’s Search Liaison Danny Sullivan has said plainly: author bios aren’t a ranking signal. Add one because it’s honest, not for the promised lift.
SE Ranking tracked what happened when Gemini 3 became the default for AI Overviews in January: roughly 42% of previously-cited domains got dropped overnight. The sites didn’t change. The model did.
A citation you have today isn’t one you’re guaranteed to keep.
Check out my latest article: Which Browser Extensions Show SEO + AI Scores? linkedin.com/pulse/which-bro… via @LinkedIn
Check out my latest article: What US SMBs Need in Audit Automation linkedin.com/pulse/what-us-s… via @LinkedIn
Ahrefs tracked what Grok actually cites, despite living inside X: Reddit (16.3%), YouTube (15.1%) and Facebook (13.9%) make up 45% of its citations combined.
Built into X doesn’t mean sourced from X. A “post more on X” Grok strategy is aimed at the wrong platform.
BrightEdge tracked video citations across ChatGPT, Perplexity, and Google’s AI results: YouTube is cited ~200x more than any other video platform. Vimeo and TikTok each sit near 0.1%.
If your AI-visibility video plan is TikTok-first, that’s not where it’s landing.
Perplexity’s Comet browser went free in October 2025. By Q1 2026 it had over 3 million monthly users.
AI search isn’t staying inside a chat window — it’s becoming the browser itself. Worth knowing where your site stands either way.
Search Engine Land audited nearly 2 million sessions across 15 domains: the passages ChatGPT actually quotes contain zero links more than 91% of the time.
Loading your key answer with outbound links may make it less quotable, not more credible.
51% of B2B software buyers now start research in an AI chatbot before Google, up from 29% a year ago, per G2.
Before you panic: WhatConverts found 85% of them still cross-check through traditional search before actually buying. AI starts the search. It rarely ends it.
The “41% vs 15%” FAQ-schema stat gets cited as proof schema boosts AI citations. Source: one vendor’s uncontrolled 50-site study. Ahrefs ran the controlled version on 1,885 pages: the lift was near zero.
A stat with no controlled study behind it is marketing, not evidence.
Claude’s citations don’t look like ChatGPT’s. Per Muck Rack: only 36% of Claude’s journalism citations are under a year old, vs 56% for ChatGPT — Claude leans toward the NYT, Atlantic, Economist.
One GEO playbook for “AI search” is actually three different playbooks.
Is GPTBot blocking your website? Check this #robotstxt #Shorts #aeo #cha... youtu.be/SNBVlHDtlFw?si=1Oz6… via @YouTube
Searchable ran 72,000 prompts about UK retailers across Gemini, Perplexity and ChatGPT: 64% had at least one factual error — bad locations, dead links, wrong businesses.
Pattern in their broader data: the smaller the business, the more AI gets wrong. Free check: 👇
Similarweb analyzed ChatGPT’s answers on US desktop: only 6.8% cite an external source at all. Travel tops out at 22.6%, education bottoms out at 4.8%.
Most ChatGPT answers aren’t citing anyone — they’re just answering from what the model already knows.
Semrush found something unexpected: Wikipedia’s share of ChatGPT’s answers fell from ~55% to under 20% in a few months. Perplexity and Google’s AI Mode didn’t move the same way.
“Get cited on Wikipedia” isn’t a stable GEO bet if ChatGPT itself is drifting away from it.