Source Tracking

🔗 Source Tracking

Visibility tells you where you stand. Sources tell you why — and exactly what to do about it.

When your Visibility Score is 66%, that's useful information. But it doesn't tell you why it's 66% and not 90%. Source tracking answers that question. By capturing every web page that AI models cite in their responses, AICarma transforms a passive score into a concrete action plan.

The Bottom Line

Sources are the bridge between knowing your score and improving it. They turn AI visibility from a monitoring exercise into a content optimization strategy with clear, actionable next steps.

How Source Tracking Works

Every time AICarma queries an AI model, we capture not just the response text (for Visibility, Sentiment, and Position scoring) but also every web source URL that the model cites or references.

This happens automatically across all models in your plan. Models that ground their responses in web sources — particularly Perplexity — provide the richest source data, but even models like ChatGPT and Gemini increasingly cite web content.

Every response is analyzed — source URLs are extracted alongside your core metrics
Sources are linked to prompts — you can see which sources appeared for which questions
Your brand and competitors are tracked — the same source analysis runs for every brand you monitor

Sources vs. Citations — What's the Difference?

Not every source that an AI model consults ends up being explicitly cited in its answer. Understanding this distinction is key to knowing where to focus your optimization efforts.

🔗 Sources

Web pages the AI model consulted when researching its answer
The full pool of content that shaped the response
Includes pages used for background knowledge, fact-checking, or context
A source may influence the answer without being named

📌 Citations

The subset of sources the model explicitly referenced in its answer
Directly linked or named in the response the user sees
These are the pages that visibly shape how AI presents brands
Citations are proof of authority — AI trusts this content enough to name it

Why This Matters for Your Strategy

Think of it this way: a model might read 20 web pages to research "best CRM tools" but only cite 5 of them in its answer. If your competitor's page is among those 5 citations, their brand gets direct visibility. If your page was only a background source, it helped inform the answer but didn't earn your brand a mention.

This is where the actionable insight lives:

Cited sources → direct visibility. If a page cites your brand, it directly boosts your Visibility and Position scores. Optimize for getting cited, not just read.
Uncited sources → indirect influence. Your content may shape the AI's opinion without earning a citation. This still matters for Sentiment, but it won't move your Visibility score.
The goal is citations. Create content so authoritative, comprehensive, and well-structured that AI models don't just read it — they reference it by name.

Three Types of Source Insight

The source URLs captured by AICarma fall into three categories, each with its own strategic value:

🏠 Your Pages

Web pages from your domain that AI models cite in their responses.

What this tells you:

Authority signals — which of your pages AI models trust enough to reference
Content that works — your cited pages reveal the format, depth, and topics that AI models prefer
Missing pages — if certain prompts never cite your content, you've found your content gaps

🎯 Competitor Pages

Web pages from competitor domains that AI models cite when answering the same prompts.

What this tells you:

What you're losing to — the specific competitor content that AI prefers over yours
Their content strategy — blog posts, comparison pages, case studies, or documentation that earns citations
Gaps to close — if a competitor is cited for a prompt where you're not, you know exactly what to write

🌐 Third-Party Sources

Independent websites — review sites, industry publications, forums, directories — that AI references.

What this tells you:

Where AI gets its information — the ecosystem of trusted content that shapes AI answers in your industry
Outreach targets — sites that AI trusts are high-value targets for PR, partnerships, guest posts, or listings
Reputation signals — review sites or forums cited by AI are actively influencing how AI describes your brand

From Insight to Action

Source tracking is powerful because it turns every insight into a specific action. Here's how to use it:

When Your Pages Are NOT Being Cited

This means AI models don't find your content authoritative enough for the topic — or it doesn't exist yet.

Create the missing content. If a competitor is cited for "best CRM for small business" and you're not, write an in-depth article or landing page targeting that exact topic.
Improve existing content. Look at the competitor pages being cited — study their structure, depth, and format. Then make yours better.
Optimize for AI readability. Ensure your pages have clear headings, structured data, and concise answers that AI models can easily extract.

When Your Pages ARE Being Cited

Reinforce what's working.

Double down on winning content. Your cited pages are the format and topics that AI trusts — create more content in the same style.
Keep content fresh. AI models often prefer up-to-date content. Regular updates to your most-cited pages maintain their authority.
Track changes over time. If a previously cited page stops appearing, investigate what changed — your content, a competitor's content, or the AI model itself.

When Third-Party Sources Dominate

These are the independent voices that AI trusts in your industry.

Review sites (G2, Capterra, TrustRadius) — ensure your profiles are complete, updated, and have strong reviews.
Industry publications — pursue PR outreach, contributed articles, or expert quotes to increase your presence.
Community platforms (Reddit, forums, Stack Overflow) — engage authentically; AI models heavily weight genuine community discussions.
Directories and comparisons — make sure you're listed and accurately described on comparison sites AI trusts.

Perplexity — The Richest Source Data

Among all AI models, Perplexity provides the most comprehensive source data. Unlike models that sometimes generate answers from training data alone, Perplexity grounds every answer in live web sources — and explicitly cites them.

This makes Perplexity source data uniquely valuable:

Real-time web citations — every Perplexity response comes with source URLs, giving you a direct view of what the web-search-augmented AI trusts right now
Content freshness signal — Perplexity actively searches the web, so its sources reflect current content authority rather than stale training data
Competitive landscape in real time — see which competitor pages are winning the web search + AI combination that's quickly becoming the default way people discover brands

All Models, One Source Stream

AICarma captures sources from every model that provides them — not just Perplexity. As more AI models adopt web-grounded responses (like ChatGPT with browsing and Gemini with Google Search), your source data becomes richer across the board.

See Which Pages AI Trusts

Start tracking the web sources behind your AI visibility scores — and turn passive metrics into an actionable content roadmap.

Start Your Free Trial