Understanding Sources
Sources are the websites AI engines cite when answering your prompts. Tracking them reveals which content influences AI knowledge about your brand and category.
Overview
The Sources page shows you which external websites AI engines cited when generating responses to your prompts. This is one of the most actionable data points in AI Brand Report — because the sources AI cites are a direct signal of which content shapes its knowledge.
Note: Citation data is currently available for Perplexity responses. Other engines do not expose citations in their API responses.
Why Sources Matter
AI engines like Perplexity don't just generate text from memory — they actively retrieve and cite sources in real time. The sources they choose to cite determine:
- What facts they present — if a cited source describes your brand inaccurately, that inaccuracy may appear in AI responses
- Which competitors they mention — if a comparison article cited by Perplexity favours a competitor, that bias will appear in the response
- The overall framing — review sites, news articles, and product directories each frame brands differently
Getting your brand featured on sources that AI engines trust and cite is one of the most direct ways to improve your AI visibility.
What You'll See
Citation List
A list of all URLs cited by AI engines across your prompt runs, showing:
| Column | Meaning |
|---|---|
| Source URL | The specific page that was cited |
| Domain | The root domain the URL belongs to |
| Times cited | How many prompt runs referenced this source |
| Associated prompts | Which of your prompts triggered this source to be cited |
Domain Summary
Sources grouped by domain — useful for spotting which websites AI most frequently draws on when discussing your category. High-frequency domains are the ones most worth targeting with PR, guest content, or link-building efforts.
Types of Sources
Understanding the type of source helps you prioritise which to target:
| Source type | Examples | Why it matters |
|---|---|---|
| Review platforms | G2, Capterra, Trustpilot | Often cited for product comparisons and sentiment |
| Industry publications | TechCrunch, Forbes, niche trade press | Shape narrative and brand positioning |
| Directories & listicles | "Best X tools" roundups | High citation rate for category queries |
| Official websites | Your own domain, competitor domains | Domain citations signal direct brand knowledge |
| Documentation & guides | Help centres, how-to articles | Cited for feature and capability questions |
| News articles | Press releases, news stories | Influence recency and topical relevance |
Acting on Source Data
Sources that mention competitors but not you
If a frequently-cited source compares tools in your category but doesn't mention your brand, that source is actively working against your AI visibility. Prioritise:
- Getting listed on comparison sites or directories that AI trusts
- Reaching out to the author or publication to be included in future updates
- Creating competing content that AI may prefer to cite
Sources that cite you negatively
If a source with a negative review or critical article is being cited frequently, AI responses about your brand will reflect that negativity. Consider:
- Responding publicly to reviews to add context
- Building more positive coverage elsewhere to dilute the negative signal
- Checking whether the cited content is outdated and requesting updates
Your own domain as a source
If your own website is being cited, it means AI engines consider your content authoritative on specific topics. Strengthen this by:
- Publishing comprehensive guides and long-form content on your key topics
- Ensuring your content is well-structured, factually accurate, and regularly updated
- Using schema markup so your content is easy for AI crawlers to parse
Tips for Improving Your Source Profile
- Audit your current sources — identify which domains are being cited and whether they portray your brand accurately
- Target high-frequency domains — focus PR and content efforts on sites that AI already cites for your category
- Build diverse coverage — AI engines weight a variety of sources; don't rely on a single domain
- Keep content fresh — recency signals matter; outdated content may be deprioritised in favour of newer sources
- Monitor changes — re-run prompts after publishing or securing new coverage to see whether citation patterns shift