> ## Documentation Index
> Fetch the complete documentation index at: https://docs.peakmark.cc/llms.txt
> Use this file to discover all available pages before exploring further.
# Sources & Citations: What AI Reads Before Answering
> Peakmark surfaces the web pages, publications, and platforms that AI models cite when recommending brands in your category — your citation roadmap.
Before an AI model recommends a brand, it reads. It reads blog posts, listicles, review profiles, Reddit threads, industry publications, and comparison pages — and the brands that appear in those sources are the brands that end up in the answer. Peakmark captures the citations that AI models surface when answering category prompts and organizes them into a dashboard that shows you exactly which sources drive recommendations in your space. This turns "we need more coverage" from a vague directive into a prioritized, specific list of target publications — the precise pages and domains you need to earn placements on to become the brand AI recommends.
## What sources and citations means
Many AI models — particularly Perplexity and the web-browsing modes of ChatGPT — explicitly cite the web pages they consulted when forming their answer. These citations are the documented evidence of which content AI trusts enough to quote. When Peakmark runs your daily prompt set, it captures every citation that appears in the responses and attributes it to the prompt, the model, and the brand associated with that answer.
The result is a growing, daily-updated map of the sources AI uses to form opinions about your category.
## Why citations are the lever
The data is unambiguous: **85.5% of AI answer citations come from earned media** — sources you don't own or control, from third-party publications, review platforms, and community discussions.
This means the content on your own website, however well-optimized, is responsible for a small fraction of what AI reads before it makes a recommendation. The majority of AI's information about your brand comes from what others have written about you. Knowing which external sources AI trusts in your category tells you exactly where to focus your digital PR, community, and content efforts.
Of AI answer citations come from earned media — third-party sources you need to earn, not own.
Google's top results and ChatGPT's citations overlap by only around 10%. SEO rankings don't predict AI citations.
Answer engines update quickly. New coverage you earn can begin appearing in AI citations within weeks.
## What the citations dashboard shows
### Domain and page list
The citations dashboard lists every domain and specific URL that AI models cited when answering category prompts. Each entry shows how often that source was cited across all prompts and models, so you can see at a glance whether a domain appears once or dozens of times.
### Brand attribution
For each citation, Peakmark records which brand was associated with the AI response it came from. This tells you whether a source is being cited when AI recommends your brand, a competitor, or neither. A high-authority domain that consistently appears alongside competitor mentions — but never yours — is a citation gap you can act on.
### Citation frequency
Sort the citations list by frequency to see the most-cited domains in your category. These are the publications that collectively form AI's frame of reference for your space. The top 10 domains on this list are your highest-priority earned media targets.
### Model-by-model breakdown
Different AI models draw from different pools of sources. Perplexity tends to cite recent news and review content heavily. Gemini draws from the broader Google index with its own weighting. ChatGPT in web mode has its own citation patterns. The model breakdown in the citations dashboard shows you which sources matter most on which models — so you can match your earned media strategy to the specific model where you most need to improve.
| Source type | Typically cited by | Why it matters |
| ------------------------------------------- | ------------------- | --------------------------------------------------------------- |
| Review platforms (G2, Capterra, Trustpilot) | All models | AI checks reviewer consensus before recommending products |
| Industry listicles and roundups | ChatGPT, Perplexity | "Best X tools" pages are among the most-cited content formats |
| Reddit threads | All models | Reddit is among the most-cited domains in AI search overall |
| Tech and trade press | Perplexity, Gemini | Earned press coverage signals authority to all major models |
| Comparison and alternative pages | ChatGPT, Perplexity | Head-to-head comparisons frequently surface in category queries |
## How to use the citations data
### Build your target publication list
Sort the citations dashboard by frequency and export the top 10–20 domains. These are the publications AI has already demonstrated it trusts in your category. Share this list with your PR or content team as the starting point for outreach — not a general "get more coverage" brief, but a specific list of named targets with evidence of AI relevance.
### Find citation gaps
Filter the citations list to show sources that are frequently cited alongside competitor mentions but not your own. Each of these is a specific gap: a domain AI trusts that it hasn't yet associated with your brand. A placement on that domain is a direct, traceable GEO action.
### Track new coverage as a leading indicator
When you earn a new press mention, a listicle placement, or a Reddit thread appearance, watch for it in the citations dashboard in the days and weeks that follow. New citations are your leading indicator of future visibility gains — they typically show up in AI answers before you see the movement in your visibility score.
Share the citations list with your PR or content team. It turns "we need more coverage" into a prioritized list of specific target publications.
## From citations to action
Understanding which sources AI cites is only valuable if you can act on it. Peakmark's managed service builds the earned media presence — digital PR placements, community participation, review profiles — that turns citation gaps into citations won.
Our team secures placements in the listicles, press roundups, and community threads that AI already cites in your category.