The sources behind every AI answer
Engines build answers from pages they trust. See which ones they cite for your prompts, what kind of pages they are, who wrote them, and which ones leave you out.
- All eight AI engines on every paid prompt run
- A 95% interval on visibility and citation rates
- Every run priced before it starts
Opening soon: sign-up opens with live measurement, after our provider pilot.
- Audio reviewsReviews and directories21%
- jbl.comYour site16%
- Audio communityForum13%
- sony.comCompetitor site11%
- apple.comCompetitor site8.9%
- Travel publicationOther6.8%
RookAtlas's citation tracker records every source each answer shows, groups them by site and by kind of page (reviews and directories, marketplaces, news, Wikipedia, Reddit, YouTube, forums, your own site and your competitors'), reads the most cited pages for their title, author and date, and flags the sites that cite competitors but not you.
Kinds of pages, not just domains
Knowing an engine leans on review sites or Reddit tells you where to show up. Every cited page gets a type from rules over its address, so each label can be explained.
- Reviews and directories, marketplaces, news, Wikipedia, Reddit, YouTube and forums
- Your own pages and your competitors' kept apart
- A filter, a label and a breakdown per type
- Reviews and directories25%
- Competitor site24%
- Forum17%
- Your site16%
- Other14%
- Reference3.4%
Labels come from rules over each page's address, so every one can be explained. Filter Sources by any type.
Authors you can write to
A byline is a person. RookAtlas reads the most cited third-party pages for their title, author, site and date, politely and within robots.txt, and lists who writes the pages engines trust.
- Authors of cited third-party pages
- Publication and update dates
- Shared by every workspace, read at most once a month
- The best headphones for long flightsMaya Lind · Travel publication · Updated Sep 14, 202610 answers
- Wireless headphones compared: 2026 editionTheo Okafor · Audio reviews · Published Aug 2, 202637 answers
- What we actually wear on long runsAna Ruiz · Running community · Published Jun 21, 20267 answers
- Leaving AirPods without losing your setupNo byline · iPhone guide · Updated Mar 3, 20265 answers
Source gaps: cited for them, not for you
A site cited in answers that name a competitor but not you is a source gap. It shows your share among the answers citing it next to your share overall, and becomes an outreach target in one click.
- At least three answers before a site counts
- Your share among citing answers and overall
- Graded correlation: branded mentions go with visibility
Travel publication
Shown as a source in 6 answers that name a competitor but not JBL.
How it works
- Step 1
Run your prompts
Every source an answer shows is kept with it.
- Step 2
See who's cited
Sites, pages, page types and authors, with your owned citation rate.
- Step 3
Close the gaps
Source gaps feed outreach; your own weak pages feed the content studio.
What's inside
Owned citation rate
How often answers cite a page on your own domains.
Citation share by site
Each site's share of all citations, counted once per answer.
Page types
Rule-based labels for every cited address, explainable one by one.
Authors and dates
Bylines and dates of the most cited third-party pages.
Source gaps
Sites that cite competitors but not you, ranked.
Retrieved, not cited
Pages an engine read but didn't cite, where it shows both.
What it can't do
Engines show sources differently: some list every page they read, some only the ones they quote, and an answer can name you without citing anything. Authors come from the page's own byline, read politely and only for the most cited pages.
Works with
All featuresQuestions
Prices, credits and plans are on the pricing page .
What's the difference between a mention and a citation?
A mention is your brand named in the answer's text. A citation is a page the answer shows as a source. You can be cited without being named, or named with no page of yours cited, and each calls for a different fix.
How are page types decided?
By rules over the page's address: known review sites, marketplaces, news sites and communities, plus your own and your competitors' domains. Rules keep every label explainable and need no model calls.
Does RookAtlas read pages that block crawlers?
No. It obeys robots.txt (its own tokens, else the rules for everyone), identifies itself honestly and reads only the most cited pages.