The scoring model
The 14 checks behind an AI citation score
Every check has a page, a weight and a reason: 25 points decide whether an engine can retrieve the page at all, 15 whether it reads as a source worth trusting, and 60 whether there is a passage worth quoting.
Score my page- 14 checks published
- 100 points, all accounted for
By Iliass El Barhoumi · Published · Last updated
Retrieved 25 points
The Retrieved lever is worth 25 points and asks: Can an engine fetch this page at all?
Ranked 15 points
The Ranked lever is worth 15 points and asks: Does this read as a source worth trusting?
Selected 60 points
The Selected lever is worth 60 points and asks: Is there a passage worth quoting?
- Answer-first structure13 ptsWhat is answer-first writing?
- Chunk extractability10 ptsWhat makes a passage extractable by an AI engine?
- Heading hierarchy9 ptsDo headings matter for AI search?
- Scannable formatting9 ptsDoes formatting change whether AI quotes your page?
- Citable specifics8 ptsWhy do AI engines prefer pages with numbers in them?
- Q&A structure7 ptsDoes an FAQ section help you get cited?
- Structured data4 ptsWhich schema types help with AI citation?
Frequently asked questions
How is the CrawlRadar score calculated?
It is the sum of 14 checks, each with a published weight: 25 points for whether an engine can retrieve the page, 15 for whether it reads as a trustworthy source, and 60 for whether it has a passage worth quoting.
Which check matters most?
Answer-first structure, at 13 points: whether each section answers its heading in its first sentence. A blocked search crawler or a noindex tag costs fewer points but is flagged critical, because it makes the page uncitable outright.
Does a high score guarantee AI citations?
No. The score measures eligibility: whether an engine can retrieve, trust and quote the page. Whether it is cited for a given query also depends on the question and the competing sources, which no tool can observe from outside.