Platform · External Sources
Map the sources that feed the answer
You do not optimise the model: you optimise the sources it learns from. This view tells you which they are.
The brands we measure every day
The metrics
From the domain to the link, from the link to the prompt
You do not optimise the model: you optimise the sources it learns from. Every citation is attributed to a domain, classified by type and traced back to the question that produced it.
-
Domains by category
Thousands of domains feed the answers in your market. The first reading is the mix: how much weight brands, retailers, publishers and user-generated content carry.
Domains by category all models3.758 domains- Brand and corporate 31%
- E-commerce 25%
- UGC 19%
- Media and publishers 18%
- Institutional and reference 7%
-
Citations by source type
The number of domains is not the number of citations: a few categories concentrate the volume, and they are almost never your own pages.
Citations by source type last 90 days
-
The domains that bring the most citations
Type, domain authority, cited links and distinct prompts: this is the list the earned media plan comes from.
Top referring domains links · domain authority- media-a.com 23.539 DA 54
- yourbrand.com 8.710 DA 44
- video-ugc.com 7.165 DA 100
- community-a.com 2.709 DA 82
-
From the domain to the single URL
A single page can generate hundreds of citations for an entire domain: it is the shift from “hold a site” to “hold a page”.
Drill-down: domain → link → prompt media-a.com- /category-guide 289× 5 prompts
- /comparison-2026 228× 5 prompts
- /how-to-choose 190× 4 prompts
-
Who gets cited on a single question
Open a prompt and you see the domains cited that day, by model: it is the proof of which sources built that answer.
Domains cited per prompt 16 citations · 11 domains- yourbrand.com · brand 3× Brand
- media-a.com · media 3× Media
- retailer-b.com · e-commerce 1× E-comm
-
The effect of publishing is visible
A new page in the corpus enters the citations within weeks: attribution becomes verifiable rather than claimed.
Citations of your domain last 90 days88% da 12%- -90g
- -60g
- -30g
- today
Generative engines systematically favour third-party sources over brand-owned content: seeing which ones is half the work.
Frequently asked questions
Taken verbatim from the monitored prompts. FAQPage schema.
How do you classify sources?
With a six-category taxonomy applied to the domain: brand, e-commerce, media, UGC, institutional, reference. Revisable per project.
Can I see competitors’ sources?
Yes. It is often the most operational figure in the whole platform.
Is the mock-up data real?
No: domains and values are fictitious. Customer data is never used as an example.
Can it be exported?
Yes, as CSV and via API on the plans that include it.
Try it on your own data
Start the 7-day free trial yourself (card required, no call needed): see your real data before choosing a plan.
