What is AEO?
AEO (Answer Engine Optimization) is making a page easy for a system that answers questions directly — ChatGPT, Perplexity, Google's AI Overviews — to read, quote and attribute.
It rewards structure over volume: a question as a heading, a complete answer in the first two sentences, facts that stand on their own without the surrounding page, and machine-readable markup. A page can rank well in classic search and still be useless to an answer engine, because the answer is buried in paragraph nine.
What is GEO?
GEO (Generative Engine Optimization) is about being cited as a source when a model generates an answer, rather than merely being crawlable.
Citation depends on things classic SEO never measured: whether your claims are attributable to a named author, whether your data is stated in a form that can be quoted, whether your entity is linked to the wider web through sameAs references, and whether your content is fresh enough for a model to trust it.
Does this replace SEO?
No. Classic SEO is still the foundation, and most AI visibility problems are ordinary SEO problems wearing a new name.
Crawlability, canonical tags, internal linking, page titles, structured data and speed all still decide whether you are indexed at all — and answer engines lean heavily on that same index. ProvenVisible scores SEO, AEO and GEO side by side so you can see when a change helps one and costs another.
What is llms.txt?
llms.txt is a proposed plain-text file at your site root that tells language models which pages matter and what your site is about.
It is a convention, not a standard, and Google has said it does not use it. That is why ProvenVisible scores it as a low-weight signal: worth having, not worth reorganising your site for. ProvenVisible can generate the file deterministically from your own pages.
Which AI crawlers should I allow?
It depends on whether the crawler feeds an answer engine that can cite you, or only collects training data.
These are different trades. A crawler that powers live answers can send you attribution and traffic; a training-only crawler takes the content with no return path. ProvenVisible recognises 26 of these agents and separates them into those two groups, so your robots.txt decision is an informed one.
How is AI visibility measured?
By asking the engines the questions your buyers actually ask, then recording whether you appear, where, and how you are described.
That produces share of voice against named competitors, a citation count, and sentiment. It is sampling, not a census — so the honest reading is a trend over time, not a single number on a single day.
Why did my score go down after an update?
Usually because coverage grew: new checks found real problems that were always there but were not being measured.
A score is only comparable when the method behind it is fixed. ProvenVisible keeps the scoring deterministic and records which checks ran, so a drop can be traced to either your site changing or the method changing — never to a model having a different mood.