Filmfolk ranked #1 on Google for "corporate videographers London." Strong backlink profile. Well-optimized content. 85+ domain authority. By every conventional SEO metric, they were winning.
When we tested their AI presence, they were cited in zero of 63 tracked prompts across ChatGPT, Claude, Perplexity, and Gemini. Not once.
This isn't an edge case. It's the norm. And it reveals a fundamental truth most marketers haven't internalized: Google rankings and AI citations are different retrieval systems with different signals.
Two Systems, Two Goals
Google is a document ranking system. It crawls pages, indexes content, evaluates authority signals (backlinks, E-E-A-T, technical quality), and returns a ranked list of URLs. The goal: surface the most relevant document for a query.
AI language models with retrieval capabilities — ChatGPT, Claude, Perplexity, Gemini — work differently. They don't return documents. They generate answers. And to generate accurate answers, they retrieve from a knowledge layer structured around entities, not pages.
Key insight: Google ranks pages. AI systems retrieve entities. If your brand isn't structured as a clear, trusted entity with consistent signals across weighted sources, you won't be cited — regardless of your Google ranking.
What Google Cares About vs What AI Cares About
| Dimension | Google (SEO) | AI Systems (GEO) |
|---|---|---|
| Goal | Page rankings, clicks | Entity citations, brand mentions |
| Unit | Pages, keywords | Entities, relationships |
| Primary Signal | Backlinks, keyword relevance | Entity confidence, source trust |
| Content Format | HTML pages, text content | Structured data (JSON-LD, llms.txt) |
| Authority Measure | Domain authority, backlinks | Source trust hierarchy (owned, earned, expert) |
| Freshness | Crawl frequency, publish date | Signal decay, continuous reinforcement |
| Success Metric | Rankings, traffic, CTR | Citation rate, prompt coverage |
Why SEO Signals Don't Transfer
1. Backlinks ≠ Entity Confidence
Google weighs backlinks heavily. A link from a high-authority site signals trust and relevance. But AI systems don't evaluate backlinks the same way. They evaluate entity confidence — how clearly and consistently your brand-role-category relationship is represented across sources.
You can have 10,000 backlinks and still have weak entity confidence if those sources don't consistently describe who you are, what you do, and why you're relevant.
2. Keyword Optimization ≠ Retrieval Priority
SEO optimizes for keyword matching. If someone searches "corporate videographers London," Google looks for pages with those keywords in titles, headers, and body text.
AI systems don't match keywords — they retrieve entities based on semantic understanding. When someone asks ChatGPT "who are the best corporate video agencies in London," the model retrieves brands it recognizes as entities in that category, not pages with keyword matches.
3. Content Volume ≠ Source Trust
SEO rewards content volume. More pages, more keywords, more opportunities to rank. But AI systems weight sources by trust hierarchy, not volume.
A single mention in an industry report (expert source) carries more weight than 100 blog posts (owned source). Most brands have heavy owned content but weak signals in higher-trust tiers — which is why they don't get cited.
4. Domain Authority ≠ Entity Authority
Google's domain authority measures link equity across your entire site. AI systems evaluate entity authority — how credible your brand is within a specific category based on proof points, client outcomes, and third-party validation.
A high-DA site with vague service descriptions and no verifiable outcomes has low entity authority. A lower-DA site with clear role definition and specific proof points has high entity authority.
The Filmfolk Case: Strong SEO, Zero AI Presence
Filmfolk is the perfect case study. Before working with RAG Signal:
- Google ranking: #1 for "corporate videographers London"
- Domain authority: 85+
- Backlinks: 2,400+ from quality sources
- Content: 50+ optimized pages
- AI citation rate: 0% (0 of 63 prompts)
Why the disconnect? Their SEO was strong, but their entity signals were weak:
- No structured data (JSON-LD schema)
- Inconsistent brand-role-category signals across pages
- Heavy owned content, minimal earned or expert sources
- No llms.txt or machine-readable entity data
- Vague service descriptions, no specific proof points
After a 90-day Brand Memory build focused on entity clarity and source trust, Filmfolk achieved 81% citation rate — cited in 51 of 63 prompts. Their Google rankings didn't change. Their AI presence transformed.
Result: Same Google ranking. 0% → 81% AI citation rate. Read the full case study →
What AI Systems Actually Look For
If backlinks and keywords don't drive AI citations, what does? Three core signals:
1. Entity Clarity
AI systems need to understand who you are, what you do, and what category you belong to. This requires:
- Consistent brand-role-category signals across all sources
- Structured data (JSON-LD schema markup, llms.txt)
- Clear entity relationships (founder, location, clients)
- Machine-readable formats AI can parse and retrieve
2. Source Trust Hierarchy
AI systems weight sources by trust level:
- Owned sources (your website) — foundational but lowest trust
- Earned sources (press, reviews) — higher trust
- Structured sources (schema, knowledge graphs) — high priority
- Expert sources (academic, industry reports) — highest trust
Most brands optimize owned sources (SEO) but neglect higher-trust tiers (GEO).
3. Factual Proof Points
AI systems prefer specific, verifiable facts over vague claims:
- "81% citation rate in 90 days" (specific, measurable)
- "Filmfolk: 0% to 81% AI citation rate" (verifiable outcome)
- "Five-step Adaptive RAG framework" (concrete methodology)
- "Ph.D. researcher at Istanbul Bilgi University" (authority signal)
These proof points must be consistent across all sources. Conflicting information reduces entity confidence and citation likelihood.
SEO vs GEO: Complementary, Not Competing
This isn't an argument against SEO. SEO and GEO (Generative Engine Optimization) are complementary disciplines:
- SEO drives traffic from search engines (Google, Bing)
- GEO drives citations from AI systems (ChatGPT, Claude, Perplexity)
Both matter. But as buyers increasingly research vendors through AI interfaces, GEO becomes the determining factor in whether your brand makes the shortlist.
The mistake is assuming SEO work automatically translates to AI visibility. It doesn't. The signals are different. The systems are different. The optimization strategies must be different.
What This Means for Your Brand
If you have strong Google rankings but aren't being cited in AI answers, the problem isn't your SEO. The problem is your entity signals.
You need to:
- Audit your entity clarity (brand-role-category consistency)
- Map your source trust profile (owned vs earned vs expert)
- Implement structured data (JSON-LD, llms.txt)
- Build factual proof points (specific, verifiable outcomes)
- Measure citation rate (not just rankings)
This is what we do in the AI Presence Audit — test real buyer prompts across all major AI models and show you exactly where your entity gaps are.
Find out where you stand
Free AI presence audit. We'll test real buyer prompts and show you exactly where your entity signals are weak.
Get your audit →