We got tired of guessing. So we built a way to know.
Cited exists because the GEO industry was giving advice nobody had tested. We decided to test it.
The problem that wouldn't let us sleep.
Here is what bothered Brian de Souza, our founder. AI engines were changing how people find information, but the advice companies were getting was mostly untested. "Add schema markup." "Write longer content." "Use more headings." Reasonable ideas, but nobody had measured whether they worked.
We started by asking a simple question: which content signals actually make AI engines cite a page? Not "which signals sound like they should help," but which ones measurably increase citation rates across ChatGPT, Gemini, Claude, and Google AI Overviews?
To answer that, we built a research pipeline. In 2026, we collected AI responses to real queries, matched them to cited pages, and measured 290+ validated signals on each page. Then we ran logistic regression, odds ratios, and multiple-testing correction. Academic rigor, applied to a business problem nobody else was measuring.
What Brian and the team found surprised us. Some of the most popular GEO advice had zero measurable effect. And some of the most impactful signals were things almost nobody was talking about. That gap between popular advice and measured reality became the foundation for Cited.
Founded by Brian de Souza.
Founder · Ex-Microsoft PM
Brian de Souza is a former Microsoft product manager who built Cited after GEO consultants kept pitching tactics without evidence. He came in without an SEO background, which helped. He asked basic questions instead of repeating industry assumptions. In 2026, he shipped the research pipeline that powers Cited. It tests 290+ validated GEO signals against real AI citations across four engines (Cited Research, 2026). Today he runs Cited end to end: research, product, and customer audits.
I built Cited because every GEO consultant I asked couldn’t show me the data behind their advice.
Find Brian online
Convictions that shape everything we build.
Measurement over opinion
The GEO industry runs on assumptions. We think that’s lazy and dangerous. If you can’t measure whether a tactic works, you shouldn’t recommend it. Every signal in our system has a number behind it.
Actionability over monitoring
Dashboards that show you’re invisible are interesting for about five minutes. The question that matters is: what do I do about it? That’s the question we exist to answer.
Honesty over hype
We’ll tell you when a signal doesn’t work. We’ll tell you when our data has limitations. We’d rather lose a deal by being honest than win one by overselling. GEO is real, but it’s not magic.
How we work.
Research-First
We do not recommend things we have not tested. In 2026, our research covers thousands of pages across thousands of domains and 4 AI engines (Cited Research, 2026). If a signal does not pass, it does not ship.
Engine-Specific
ChatGPT, Gemini, Claude, and Google AI Overviews each use different criteria to decide what to cite. Generic AI optimization advice ignores this. We test each engine independently and tell you exactly which signals matter where.
Implementation, not reports
We don’t hand you a PDF and disappear. We work with your team to make the changes, track the impact, and adjust when AI engines evolve. The value is in the doing, not the documenting.
30 minutes. Ranked actions. Zero obligation.
30 minutes. Real signals. Your pages.
What you get
30-minute review
A live walkthrough of the pages AI is reading and the gaps costing you citations.
Ranked next steps
Top fixes prioritized by measured lift, not generic SEO advice.
Written handoff
A clean summary your team can act on even if you never hire us.
Last updated: May 2026