Which AI platform should I optimize for first?

Microsoft Copilot is the highest ROI target - 95.29% citation rate + 60.3% top 3 placement = nearly guaranteed visibility. Follow with ChatGPT/GPT (44.7% market share, 64.85% top 3 via GPT-4o-mini) and Perplexity (30.6% market share, heavy commercial preference).

Do LLMs prefer long or short content?

It depends on the model. Claude 3.7 Sonnet averages only 1,059 characters (ultra-concise), while Google AI Overview averages 6,174 characters (highly detailed). There's no universal length - optimize based on your target platform.

Is my 2024 content still relevant for AI citations in 2025?

Yes! Our data shows that authority beats freshness. 2024 content ranks nearly as well as 2025 content. Focus on creating evergreen, authoritative content rather than constantly updating for recency.

Should I focus on awareness (TOFU) or purchase-intent (BOFU) content?

BOFU content is critical. Purchase-intent queries trigger 2.5x more competitor mentions (4.78 avg vs 1.92 for TOFU). If you're not visible when users ask "Best [category]" or "[Brand A] vs [Brand B]", you're invisible at the moment that matters most.

Why don't Claude, Mistral, and DeepSeek cite any sources?

These models use a fundamentally different trust model - they rely entirely on training data rather than real-time sourcing. They still have high citation rates (47-100% mention competitors), but provide zero attribution. To rank here, focus on becoming part of future training data through authoritative, widely-cited content.

How many competitors will be mentioned alongside my brand?

It varies dramatically by platform: Mistral: 11.69 avg (most competitive), Grok: 7.94 avg, ChatGPT/GPT: 3.54 avg (was 0.81 in Q2!), Perplexity: 1.38 avg (most selective). Position matters more than ever when 3-12 alternatives are mentioned.

Is video content (YouTube) important for GEO?

Yes, critically important. YouTube is the #1 source for Google AI Overview, #6 for Perplexity Sonar, and #9 for Gemini 2.0 Flash. If you're not creating video content, you're invisible to a significant portion of AI search traffic.

What's the biggest change from the Q2 2025 study?

ChatGPT's massive shift: Competitor mentions increased by 337% platform-wide (from 0.81 to 3.54 avg), with ChatGPT Core specifically at 490% (4.78 avg). OpenAI dramatically changed its strategy to recommend many more alternatives. Combined with 106% month-over-month growth in AI search adoption, the competitive landscape has intensified dramatically.

GEO

ChatGPT

Gemini

Perplexity

Claude

AI Search

LLM Ranking

BOFU

Citation Rate

Brand Positioning

We analyzed 184,128 queries on ChatGPT, Gemini, Perplexity and Claude. Here is what we learned.

Luca Fancello•October 26, 2025•Updated Dec 21, 2025•34 min read•

Research

The most comprehensive study ever conducted on LLM ranking factors, analyzing 184,128 queries and 1,479,145 sources across 20 AI models including ChatGPT, Gemini, Perplexity, Claude, Mistral, DeepSeek, and Grok to help you dominate AI-generated results.