Help center

How AI visibility works, how to read your results, and what to do about them.

Getting started

What AI visibility actually means — Why being recommended by ChatGPT is a different problem from ranking on Google, and what we measure instead.
Setting up your first brand — The three things to get right at setup, and the one that people usually get wrong.
Writing questions that match how customers really ask — The single highest-leverage thing you control. Most accounts get this wrong on the first try.
Reading your Overview — What each panel is telling you, and which one to act on first.

Visibility, Authority and Position — what each one means — Three numbers that sound similar and answer completely different questions.
Why other people's websites decide your AI visibility — We measured our own category and found 93.6% of citations pointed somewhere other than the brand's own site.
Why your sentiment score is probably 50 — Neutral is the normal result, not a bug — and here is what would change it.
Why your score moved (or didn't) — The usual explanations, in the order worth checking.
Why the assistants disagree with each other — Different sources, different training, different recency. Disagreement is information.

llms.txt, explained — What it is, what it is not, and whether it is worth your time.
robots.txt and AI crawlers — Which bots matter, and the one mistake that makes everything else pointless.
The schema markup worth adding — Three types, in the order they pay off — and the mistake that gets pages penalised.
Connect your AI coding tool to Kartafla — Point Claude Code, Cursor or any MCP client at your account so it can read your site readiness and apply the fixes itself.
What to do when the fix isn't on your website — The uncomfortable answer, and a practical order to work through it.

Plans, and choosing your AI assistants — What differs between plans, and how to pick your engines if your plan lets you.
How monthly checks are counted — One per brand per calendar month, shared between manual and automatic runs.
Roles and team access — Who can see what, and which role to give the person who implements the fixes.