GEO Readiness50points
How ready is your site?
This is the part you control: five signals, each worth points, each point backed by something we actually observed on your site. Fix a problem, and this number moves on the next scan.
What We Do
We Track All 5 Factors That Impact AI Search - And Give You Specific Fixes
Every point in your score comes with the evidence behind it — what we saw, on which page, on what date. No mystery number you have to take on faith. And where the evidence can't support a conclusion, we say so instead of guessing.
Two scores, kept separate
GEO Readiness vs. AI Visibility: We Measure Both
A perfectly-built website can still be missing from AI answers.
A business that gets recommended in certain categories can still have major problems under the hood keeping it from full reach.
We measure both metrics separately - so that you can understand what will give you the best improvement where it counts most - new customer acquisitions and bottom line ROI.
AI VisibilityObserved
Are you showing up?
We ask real questions on real dates and record what came back: were you mentioned, cited, and how were you described? Who got named instead? This moves more slowly, and nobody fully controls it — including us.
Your readiness score
GEO Readiness 101: How AI is ranking your online presence
Your readiness score out of 50 is made of these five. Every point requires evidence we can show you.
01
Can AI reach your site?
AI access and indexabilitySome websites accidentally tell AI systems to stay out — usually one line of settings added years ago by a template or plugin. If AI can't read your pages, nothing else on this list matters.
We checkWhether AI crawlers can reach your pages, follow your links, and read what's there.
If this is lowIt's usually the highest-priority fix on your report, and often a five-minute change.
Learn to fix this yourself02
Can AI tell who you are?
Entity clarity and site architectureAI has to understand what your business is before it can recommend it. That means your name, what you do, where you are, and how it all connects — stated clearly enough that a machine doesn't have to guess.
We checkWhether your business is plainly identified, your pages are organized sensibly, and the machine-readable facts about you are present and correct.
If this is lowYou may be getting confused with another business, or skipped because AI can't tell what you offer.
Learn to fix this yourself03
Do you answer the questions people ask?
Answer-ready content“Do you do same-day repairs?” “What does a first consult cost?” “Are you open Sunday?” Pages that answer real questions directly are easier for people and AI systems to use. Most small-business websites answer none of them.
We checkWhether your site contains direct answers to the questions customers actually ask about a business like yours.
If this is lowYou're missing the searches that matter most — the ones where someone is ready to buy.
Learn to fix this yourself04
Do you look like a real business?
Evidence and originalityReal photos. Named people. Specific details. Current information. As generating a plausible-looking website became nearly free, customers and AI systems both needed better ways to tell a real business from a generic template.
We checkWhether your pages show visible authorship, dates, supporting links, real specifics, and original content.
If this is lowYour site may read as a template with a phone number rather than a business someone can trust.
Learn to fix this yourself05
Does the rest of the internet agree with you?
External validationYour Google Business Profile, reviews, and credible mentions can reinforce the facts on your website. When public sources contradict one another, AI has a harder time knowing which version is current.
We checkWhether the story your site tells matches the story the rest of the web tells about you.
If this is lowAI has less outside evidence to confirm that your business is real, current, and operating.
Learn to fix this yourselfAI visibility
Six Engines, One Brand Truth
Your visibility results come from asking real questions across the assistants your customers use, on recorded dates, with the questions and locations written down so results can be compared over time.
- Perplexity
- ChatGPT (OpenAI)
- Claude
- Grok
- Gemini
- You.com
AI Mode and AI Overviews
We only report these when you connect your own Google Search Console and it makes the data available. We don't estimate what we can't see.
Microsoft Copilot
Microsoft Copilot is an optional connection currently in preview.
Meta AI
Meta AI is not yet included in GEO's automated scans.
One approved version of you
Brand Truth: your facts, in one place, checked for contradictions.
We keep a record of the facts you've approved about your business — your name, services, locations, leadership, and the claims you stand behind. Then we watch for contradictions: a different phone number on one page, an old service you no longer offer, or a location that changed.
Conflicting facts confuse AI systems and customers alike, and they're easy to miss on your own site.
If we find a contradiction, we show you both versions and where each one lives.The honest limits
What we won't tell you.
This is a field where certainty is easy to sell and hard to justify. Here's where we stop.
We won't promise you a spot in AI answers.
Nobody can. Answers vary by how the question is asked, the day, the person asking, and the version of the system. There is no submission form or paid placement into organic AI recommendations.
A missing citation doesn't prove you were never considered.
We can show that a page was reachable, found, cited, or mentioned. We can't prove what a system decided not to use. We'll tell you which of those we actually know.
We won't turn missing data into zero.
If a metric isn't available for your site, we label it unavailable. Missing evidence is not a bad score.
We won't take credit we can't prove.
When visibility improves after a fix, we'll show you the timing and the movement. We will not claim the fix caused it unless the evidence can support that conclusion.
Our score is ours, not theirs.
The readiness score is our own transparent measurement with published criteria. It isn't a number from inside Google, OpenAI, or anyone else.
The fine print
For the technically minded.
If you're an analyst, an agency, or just want the details, here's exactly how the evidence works.
Dated observations from the AI systems' own developer tools.
Each recorded result names the exact question asked, the model, the date, and the location. That makes results repeatable and comparable over time. It isn't a claim that every consumer app session returns an identical answer — the same question can produce different answers for different people.
Your own Google data, when you connect it.
A connected Search Console or Analytics property can show your real search and referral data. If your property doesn't expose a particular metric, we mark it unavailable rather than reporting it as zero.
Retained website evidence.
Every audit records which conditions we actually observed on which pages, and which points that evidence supports. We identify practical blockers without claiming our score is any other company's internal ranking system.
Before-and-after records.
Later scans can show that readiness, citations, referrals, or visibility changed after an approved fix. The record captures timing and movement — not proof of cause.
How questions and competitors are chosen.
The set of tracked questions is written down and versioned. It combines what you tell us about your business, the themes customers actually search, unbranded discovery questions, branded verification questions, and competitors you agree to. Fixed questions, dates, locations, models, and repeat runs are what make comparisons hold up.
Questions
How to read what you see.
What does the readiness score measure?
Fifty points across five signals: can AI reach your site, can it tell who you are, do you answer real questions, do you look like a real business, and does the rest of the internet agree with you. Every point awarded requires evidence we can show you.
Is readiness the same as visibility?
No, and we never mix them. Readiness is what we can observe about your website and your presence — the part you control. Visibility is what AI assistants actually said on specific dates when asked real questions.
Does a missing citation mean AI never saw my site?
No. We can show that a page was reachable, found, cited, or mentioned. We can't prove a system never considered it — and we won't pretend otherwise.
How do you choose the questions you track?
From what you tell us about your business, the themes your customers search, unbranded questions where you'd want to be discovered, branded questions to check what's said about you, and competitors you agree to. The set is written down and versioned so comparisons over time mean something.
What is Brand Truth?
A record of the facts you've approved about your business, checked against your site for contradictions. If we find two versions, we show you both and where each one lives.
Which AI assistants do you check?
Perplexity, ChatGPT, Claude, Grok, Gemini, and You.com. We can also use your own Google Search Console data when you connect it. Meta AI is not yet connected to GEO. The provider section above shows the status of each connection.
Start with your website
See your five signals.
Free scan. About two minutes. Read-only — no passwords, no changes to your site.