Citations
Understand where AI models learn about your brand - and how to influence what they say.
- Why citations determine what AI says about you
- How different AI models use sources differently
- The connection between citations and visibility
- What you can actually do to improve your citation profile
Your visibility score tells you whether AI mentions your brand. Citations tell you why - and more importantly, how to change it.
The source problem
Here's what most people don't realize about AI search: every response is built from sources. When ChatGPT recommends "the best running shoes for marathons," it's not making that up. It's drawing from Wirecutter reviews, Runner's World articles, Reddit threads, and thousands of other pages it has encountered.
The same is true for Perplexity, Claude, and Gemini. Different models, different training data, different real-time retrieval - but they all rely on sources.
Your visibility is a downstream effect of your citation profile.
If Nike.com appears in 67% of running shoe prompts, it's because Nike shows up in the sources those models reference. If Adidas appears in 72%, they're cited more often, on more authoritative sites, in more favorable contexts.
How AI models use sources
Not all citations are created equal. Understanding how models weight sources helps you prioritize.
Training data citations
ChatGPT, Claude, and Gemini all trained on massive datasets - essentially a snapshot of the internet. Content that existed during training is baked into the model's knowledge.
What this means for you:
- Older, established coverage persists even if articles disappear
- Getting mentioned in widely-linked content has lasting effects
- Major publications carry more weight than obscure blogs
Real-time retrieval
Perplexity is different. It searches the web in real-time for every query. SearchGPT works similarly. This means recent content matters more.
What this means for you:
- Fresh content can impact visibility immediately
- News coverage shows up faster on retrieval-based models
- You can see changes from new mentions within days, not months
The authority signal
All models weight authoritative sources more heavily. A mention on TechCrunch influences AI responses more than a mention on a random blog. Domain authority, backlinks, and content quality all factor in.
| Source Type | AI Weight | Why |
|---|---|---|
| Major publications (NYT, TechCrunch) | Very High | Widely cited, high authority |
| Review sites (G2, Wirecutter) | High | Product-focused, trusted |
| Industry publications | High | Niche authority |
| Wikipedia | Very High | Crawled extensively, treated as fact |
| Reddit/forums | Medium | Real user opinions, but variable quality |
| Company blogs | Medium | First-party content, potential bias |
| Small blogs | Low | Less reach, less authority |
Citations drive three things
1. Whether you appear at all
If no authoritative source mentions your brand in the context of "best CRM software," AI won't recommend you for that query. Simple as that.
The gap problem: When Nike checks the Heatmap and sees that Runner's World cites Adidas and New Balance but not Nike for "best trail running shoes" - that's a gap. Nike is invisible on that source for that topic. AI models drawing from Runner's World won't mention Nike.
2. How you're positioned
Being mentioned isn't enough. How you're mentioned matters.
Consider these two scenarios for the same brand:
Scenario A: "HubSpot is the leading CRM platform for growing businesses, with best-in-class marketing automation."
Scenario B: "HubSpot is an alternative to Salesforce for teams who can't afford enterprise pricing."
Both are citations. Both cause the brand to appear in AI responses. But one positions you as a leader, the other as a budget option. AI models reflect this framing.
3. Your sentiment score
Trakkr analyzes whether citations position you positively, neutrally, or negatively. Sentiment matters because:
- Negative citations can drag down overall AI perception
- Positive citations amplify when multiple sources agree
- Mixed sentiment creates hedged AI responses ("Some users report...")
The citation improvement cycle
Improving your AI visibility isn't a one-time audit. It's an ongoing cycle.
1. Discover your citation landscape
Where do you appear? Where don't you? Which sources cite competitors but not you?
This is what the Citations page shows you. Start here.
2. Analyze what's driving your scores
Look at the specific content. Are high-authority sources positioning you well? Are there gaps on sites that matter? Is there negative coverage you need to address?
3. Act on what you learn
This might mean:
- Creating content that earns citations naturally
- Reaching out to publications for reviews or inclusion
- Responding to negative coverage
- Getting listed on comparison sites
4. Measure the impact
Run research again. Did your gaps close? Did new positive citations appear? Are you winning on prompts you were losing before?
Then repeat. Brands that treat this as continuous see the best results.
What Citations shows you
The Citations page has six views, accessible via the tab bar. Use the ?view= URL parameter to link directly to any view.
Sources (default)
"Who's talking about me, and what are they saying?"
The split-pane explorer shows every website and publication that cites your brand or competitors. Click any source to see:
- Exact quotes and context
- Sentiment analysis
- Which AI models use this source
- How competitors are covered
Pages
"Which specific pages are AI models citing most?"
A flat, cross-domain page-level view. While Sources groups by domain, Pages shows individual URLs ranked by citation influence - the actual content AI draws from.
Queries
"What questions lead to citations about me?"
See the search queries that trigger citations. This reveals user intent and content opportunities:
- Discovery queries ("best project management tools")
- Comparison queries ("Notion vs Confluence")
- Problem queries ("Notion slow loading fix")
Feed
"What changed in my citation landscape?"
A chronological changelog of citation events - new pages appearing, lost citations, sentiment shifts, and competitor movements. Think of it as your citation activity feed.
Heatmap
"Where do I win vs competitors across all sources?"
Visual grid of your citation coverage vs competitors. Green means you're cited, red means competitors are cited and you're not. Patterns become obvious:
- Vertical green stripe = strong coverage
- Horizontal red row = priority gap on a single source
- Scattered red = multiple opportunities to close
Outreach
"Which gaps should I prioritize?"
Prioritized list of sources worth pursuing, ranked by potential impact. Includes contact information and submission guidelines when available.
Quick wins
Three things you can do today:
1. Find your highest-impact gap
Go to Citations → Heatmap. Filter to "Gaps only." Sort by authority. The first red cell you see is your biggest opportunity.
2. Understand your positioning
Go to Citations → Sources. Click your top-cited source. Read what they actually say about you. Is this how you want to be positioned?
3. Check competitor coverage
Pick your main competitor. How many sources cite them but not you? That number is your opportunity list.
The bigger picture
Citations connect to everything else in Trakkr:
- Dashboard visibility comes from citation coverage
- Competitor rankings reflect relative citation strength
- Prompt performance ties back to which sources cite you for which topics
- Content opportunities emerge from citation gaps
When you wonder "why is my visibility what it is?" - Citations has the answer. When you ask "how do I improve?" - Citations shows the path.
Was this helpful?