Why do some sources get cited first by AI? Trying to understand citation order
Community discussion on factors that determine citation order in AI responses. Real insights on what makes sources get cited first.
We finally got our site cited by ChatGPT and Perplexity. Great, right?
The problem: We’re always citation #3 or #4. Never first.
Why does this matter? Same as traditional search - first position gets the clicks. I’ve watched user sessions where people click the first citation and ignore the rest.
What we’ve noticed:
What we’ve tried:
What I need to know:
The difference between first and fourth citation is probably 10x the traffic.
I’ve analyzed 50K+ AI citations to understand position factors. Here’s what actually matters:
The Big Three for Citation Position:
1. Domain Authority (strongest signal)
Research shows clear thresholds:
Sites with massive backlink profiles consistently get first position. It’s the single strongest predictor.
2. Content Freshness (time decay is real)
AI systems have time decay parameters:
If you’re being outranked, check when competitors last updated.
3. Page Speed (often overlooked)
FCP (First Contentful Paint) matters:
This was surprising but consistent in the data.
Secondary factors:
Why you’re at position 3-4:
Likely strong content (gets you cited) but weaker domain authority or slower page speed than top competitors.
First position is achievable but requires winning on multiple factors, not just one.
Yes, the correlation is surprisingly strong.
Here’s the data breakdown:
| FCP Speed | Avg Citations | Position Tendency |
|---|---|---|
| Under 0.4s | 6.7 | Usually 1st-2nd |
| 0.4s-0.7s | 4.2 | Usually 2nd-3rd |
| 0.7s-1.0s | 3.1 | Usually 3rd-4th |
| Over 1.0s | 2.1 | Rarely cited |
Why speed matters for AI:
AI crawlers have timeout limits. If your page loads slowly:
Quick wins for speed:
If your competitors are at 0.4s and you’re at 0.8s, that alone could explain position 3-4 vs position 1.
It’s the lowest-hanging fruit for citation position improvement.
We moved from position 4 to position 1 for our primary category query. Here’s exactly what we did:
Starting point:
Changes made (in priority order):
Week 1-2: Technical fixes
Week 3-4: Content freshness
Week 5-6: Content structure
Week 7-8: Authority signals
Results:
| Week | Perplexity Position | ChatGPT Position |
|---|---|---|
| 0 | 4 | 3-4 |
| 4 | 3 | 3 |
| 8 | 2 | 2 |
| 12 | 1 | 1-2 |
Key insight: It’s compound effect. No single change moved us dramatically. All of them together moved us from 4 to 1 over 12 weeks.
The hard truth about first position:
Domain authority is the elephant in the room.
If your competitors have 10x your backlinks, you’re fighting an uphill battle for first position. Content and speed help, but authority is the foundation.
What this means practically:
For some queries, first position may require:
Where you CAN win first position quickly:
Strategy based on your situation:
If competitors have much higher authority:
If authority is roughly equal:
Don’t waste effort fighting for position 1 on queries where you can’t compete yet. Find winnable battles.
Adding data on what DOESN’T move position:
Things we tested that had no impact on citation order:
Things that DO move position:
The formula I use:
Position Potential =
Authority (40%) +
Speed (25%) +
Freshness (20%) +
Content Quality (15%)
If you’re losing on authority, you need to WIN on speed and freshness to compensate.
Tools that help:
We use Am I Cited to track position changes over time. When we make changes, we can see if position moves within 2-4 weeks.
Content structure matters more than people realize:
Answer-first formatting improved our position from 3 to 1 on several queries.
Before (position 3): “When considering X, there are many factors to think about. In this comprehensive guide, we’ll explore…”
After (position 1): “X works by [direct answer in 15 words]. Here’s why this matters…”
AI systems extract the first clear answer they find. If your answer is buried in paragraph 3, and competitor’s answer is in sentence 1, they win.
Structural changes that helped:
Before/After example:
Before: “Understanding AI visibility is crucial for modern businesses…”
After: “AI visibility measures how often your brand appears in AI-generated answers like ChatGPT and Perplexity.”
The second version is immediately extractable and citable. AI prefers it.
Reverse-engineer your first-position competitors:
What to analyze:
For each query where you’re position 3-4:
Create a comparison matrix:
| Factor | Position 1 | You (Position 4) |
|---|---|---|
| Domain Rating | 85 | 62 |
| Page Speed | 0.3s | 0.8s |
| Last Updated | 2 weeks ago | 6 months ago |
| Word Count | 3,200 | 2,800 |
| Expert Quotes | 4 | 0 |
| Data Points | 23 | 8 |
This matrix tells you exactly what to fix.
Reality check:
If position 1 has DR 85 and you have DR 45, content optimization alone won’t get you to first. You need a multi-quarter authority building strategy.
But if DR is similar and they’re faster/fresher, those are fixable in weeks.
This thread gave me the framework I needed. Here’s my action plan:
Immediate Analysis (This Week):
Quick Wins (Next 2 Weeks):
Medium-Term (Weeks 3-8):
Long-Term (Ongoing):
Key realizations:
New target: Move 3 queries from position 4 to position 1 within 90 days (choosing queries where authority gap is manageable).
Thanks everyone for the data-driven insights!
Get personalized help from our team. We'll respond within 24 hours.
Track where your brand appears in the citation order across AI platforms and identify opportunities to move up.
Community discussion on factors that determine citation order in AI responses. Real insights on what makes sources get cited first.
Community discussion on whether citation order matters in AI answers. Real data from marketers analyzing first-position vs. later citations in ChatGPT and Perpl...
Community discussion on how AI models decide what to cite. Real experiences from SEOs analyzing citation patterns across ChatGPT, Perplexity, and Gemini.
Cookie Consent
We use cookies to enhance your browsing experience and analyze our traffic. See our privacy policy.