GEO Audit AI search visibility 中文

When a buyer asks AI for a recommendation, are you in the answer?

Most site owners cannot see this. The engine answers from sources you do not control, never tells you, and the visitor never arrives with a referrer that says where they came from. This tool places an instrument on your site and reports back what the engines actually said.

Sharpens the buyer question we ask each engine.

Optional: brand name and facts to check

Only needed when your brand name differs from the domain.

One per line, as what AI might say => what is true. We only check claims you supply — we never guess at your facts.

Three free scans per day. Each one runs live AI queries, which cost money to produce — that is the only reason there is a limit.

Specimen reading acme-crm.example.com
ChatGPT Named you 2nd of four, described as “built for small teams”. Your pricing page cited. Mentioned
Perplexity Named a competitor four times, never you. Cited a two-year-old review you cannot edit. Absent
DeepSeek Answered from training memory without searching the web, so it measures what it already knows about you. Not compared
AI Overviews Requires Google Search infrastructure this deployment does not have. Not measured
Illustrative output, not a customer result. Every score we report is generated from a real scan of a real site.

What the scan actually checks

Every check runs in five groups — crawler access, entity clarity, answer-first structure, live prompts and technical hygiene — and they are identical in the free and full scan. What differs is the live test: the free scan asks 3 buyer questions; the full audit asks 10 and adds share of voice, source classification and AI-referral reconciliation.

Whether AI can read you at all

Fifteen crawlers tested against your robots.txt under RFC 9309, plus whether the page is readable without JavaScript. Blocking an answer crawler removes you from that engine entirely, and most owners never find out.

What the engines say

Real buyer questions asked of each configured engine, twice, with the answer analysed for whether you appeared, where in the list, how the sentence read, and who else was named.

What to change, in order

A fix list ordered by impact per hour, each item carrying the actual file content to paste — robots.txt, llms.txt, JSON-LD, an FAQ section, a comparison page.

Three rules the audit follows

These are not marketing claims about accuracy. They are constraints in the code, each covered by a test that fails if the behaviour regresses.

  1. It will not invent a score

    If your site blocks our crawler or does not resolve, the result is not scored with a reason — not a low score that implies we looked and found problems. We would rather show you a gap in our own measurement than a number we cannot defend.

  2. It will not blame you for a choice

    Blocking GPTBot is a legitimate opt-out that has no effect on whether ChatGPT cites you, so it never costs points. Only crawlers that feed answers are scored. A tool that penalises this is telling you to make a decision against your own interest.

  3. It separates knowledge from visibility

    An engine that searches the live web and misses you has a content problem you can fix. An engine answering from training memory and missing you has a knowledge problem that on-page edits will not solve. The report labels which is which instead of blending them.

Start with your own domain

The free scan takes about thirty seconds and needs nothing from you — no account, no code to install, no access to your analytics.

Run the free scan