Skip to main content

What are sources?

Discover which content shapes what AI models say about your brand

Written by Urszula Kucińska

Ever wondered why ChatGPT recommends your competitor but not you? Or why Gemini suddenly started mentioning your brand more often? The answer usually lies in the sources.

AI models rely on two types of knowledge when generating answers. The first is their built-in training data. The second is external sources, meaning web pages the model retrieves and cites in real time. These external sources are what you can actually influence, and that's exactly what Chatbeat tracks for you.

When Chatbeat queries AI models on your behalf, it collects not only the answers but also every source each model used to generate those answers. This means you can see which blog posts, review sites, comparison pages, and community discussions are actively shaping how AI talks about your category.

Why This Matters for Your Business

If a competitor's blog post keeps getting cited by ChatGPT, that page is directly influencing purchase recommendations seen by thousands of potential customers. If a G2 review page or a Reddit thread dominates Gemini's sources, it tells you that community presence matters more than your own website for AI visibility in your space.

With Sources data, you can pinpoint exactly which content drives your brand's AI visibility and where your competitors are getting their edge. You can spot content gaps and create pages that AI models will want to cite. And you can track whether your content efforts are actually translating into more citations over time.

What You'll Find in the Sources Tab

The Sources section of your project contains several views.

Key Sources. A ranked list of all URLs cited by AI models when answering your project's prompts, along with their coverage percentage and which models cited them. You can filter this view by specific AI model to see, for example, which pages ChatGPT relies on versus which ones Gemini prefers.

AI Citations. A summary of how many total citations each AI model generated across your project's prompts, along with how many unique pages were cited. This helps you understand which models rely most heavily on external sources and which ones lean more on built-in knowledge.

Citation Trends. A chart showing how citations from the top domains changed over time. This lets you spot shifts in which sources are gaining or losing influence in AI responses.

Top Citation Domains. A table of the most frequently cited domains, with their share of total citations, how that share has changed compared to the previous period, and how many of your tracked prompts they appeared in.

How to Use This Data

Sources data helps you answer practical questions. If your own domain isn't being cited, you know you need to create content that AI models can discover and reference. If a competitor's blog post ranks as a top source, you can study what makes it citable and create something better. If Reddit or G2 dominate the citations, you know that community presence and reviews matter more than your own blog for AI visibility in your space.

Combined with the rest of your Chatbeat data, Sources gives you a complete picture: not just where you rank, but why you rank there and what content is driving the results.

For tips on improving your brand's presence in AI-cited sources, check out our guide at chatbeat.com/ai-seo-rank-in-chatgpt or visit the GEO Playbook tab in your project. If you have any questions, don't hesitate to contact us.

Did this answer your question?