AI visibility score: what it measures and what it misses
What an AI visibility score actually measures, why scores from different tools disagree, and the raw data you should ask for instead of a single number.
SEO
Rank on Google
Blue links, snippets and maps — and the base layer AI reads from.
AEO
Named by assistants
Cited in the answers ChatGPT, Perplexity and Copilot give.
GEO
Synthesised by AI
The source Google AI Overviews and Gemini build their answer from.
One shared foundation: crawlable, structured, trustworthy content — run as a single program.
Most AI visibility tools give you a score. It's a convenient number, and it can be useful for spotting trends. It can also hide almost everything you need to know. Here's what goes into these scores and what to look at alongside them.
What a score is usually made of
- A set of prompts, chosen by you or by the tool.
- The assistants those prompts are run on, sometimes weighted differently.
- Whether your brand is mentioned in each answer, and sometimes where in the answer.
- Sometimes sentiment, or whether you're recommended or only mentioned.
Why scores disagree
Two tools can give the same brand very different scores, because they use different prompts, different assistants, different weights and different dates. Answers also vary from run to run. A score only means something if you know the prompts behind it and they stay the same over time.
What a score misses
- Which questions you win and which you lose. A score of 40 could be all long-tail questions or one important one.
- Who is named instead of you.
- Which sources the assistants cite, which is what you can actually change.
- Whether the mentions produce visits or enquiries.
- Whether the assistant describes you accurately.
Ask for the raw data
Whatever tool or agency you use, ask for the list of prompts, the dated answers, the competitors named and the cited sources. With those, you can see what changed and why. Without them, a rising score is hard to trust and a falling one is hard to fix.
Keep the prompt set fixed
If the prompts change every month, the score can rise or fall without anything real changing. Add new prompts as a separate set, and keep the original set for comparison.
How to read a score properly
- 1Ask how many prompts are behind it, and see the list.
- 2Check which assistants are included and how they're weighted.
- 3Look at the same score over at least six to eight weeks.
- 4Open the answers behind any big jump or drop.
Used this way, a score is a helpful summary. Used on its own, it's easy to over-read.
Our weekly reports show the raw answers behind every number. See how in our AI visibility audit.
Frequently asked questions
- What is a good AI visibility score?
- There's no universal benchmark, because scores depend on the prompts and tools used. Compare your own score over time on a fixed prompt set, and against named competitors on the same prompts.
- Why does my AI visibility score change every week?
- AI answers vary from run to run, and assistants update often. Look at trends over several weeks rather than single readings.
- What should I track instead of a score?
- Track the prompts, the dated answers, the competitors named, the sources cited, and AI-referred visits and enquiries.
AI VisibilityAEOGEO
M
Magathi
AI Visibility specialist at Edthi · SEO, AEO & GEO
Leads AI Visibility programs at Edthi, helping companies get found and recommended across Google and the AI engines. Writes about how AI search actually decides who to cite.
Ready when you are.
Book a 30-minute call, message us on WhatsApp, or drop your details and we'll reach out. We'll show you where you show up today across Google and the AI engines — and the plan to become the answer.