AI & Automation
Personas
SaaS & Startup
Time to ROI
Short-term (< 3 months)
You know what's funny? While everyone's panicking about AI "killing SEO," I've been quietly tracking what makes content appear in ChatGPT responses. And let me tell you - it's not what the SEO gurus are telling you.
Last month, I was working with a B2B SaaS client who was getting zero mentions in AI-generated responses, despite having solid traditional SEO rankings. Their competitor, with worse Google rankings, was being cited by ChatGPT constantly. That's when I realized we were playing by the wrong rules.
The truth? ChatGPT ranking signals are fundamentally different from Google's algorithm. While everyone's debating whether AI will replace search, they're missing the bigger opportunity: optimizing for AI discoverability right now.
Here's what you'll learn from my hands-on experiments:
The real signals that influence ChatGPT mentions (hint: it's not backlinks)
Why "chunk-level" thinking beats traditional page optimization
The content structure that LLMs actually prefer
How to track and measure AI mentions (because traditional analytics won't help)
The counterintuitive approach that's working for my clients
This isn't theory - it's based on real experiments with actual AI mention tracking. Let's dive into what I discovered about the new rules of the game.
Reality Check
What the SEO world is saying about ChatGPT ranking
If you've been following the SEO space lately, you've probably heard the same advice repeated everywhere. Most experts are telling you to:
Focus on "E-E-A-T" signals - The belief is that ChatGPT favors content with expertise, experience, authoritativeness, and trustworthiness. Sounds logical, right? But they're applying Google's framework to a completely different system.
Optimize for featured snippets - The thinking goes: if it works for Google's AI features, it must work for ChatGPT. So everyone's restructuring content for position zero optimization.
Build more backlinks - Traditional SEOs assume that authority signals like backlinks carry over to AI systems. They're doubling down on link building campaigns.
Create longer, comprehensive content - The "10x content" crowd believes that exhaustive, pillar-page style content will dominate AI responses.
Target question-based keywords - Since people ask ChatGPT questions, the advice is to optimize for question keywords and conversational search.
Here's why this conventional wisdom exists: It's a logical extension of what we know about Google. When faced with a new system, we naturally apply our existing frameworks. SEO professionals are experts at Google's algorithm, so they're mapping those same principles onto AI systems.
The problem? ChatGPT doesn't work like Google. It doesn't crawl and rank pages - it synthesizes information from training data. It doesn't consider backlinks or domain authority. And most importantly, it's not trying to send you to the "best" page - it's trying to give you the "best answer."
This fundamental misunderstanding is why most SEO advice for AI optimization falls flat. We need a completely different approach.
Consider me as your business complice.
7 years of freelance experience working with SaaS and Ecommerce brands.
The wake-up call came during a project with an ecommerce client. Despite having strong Google rankings for product-related keywords, they were getting zero mentions when people asked ChatGPT about their industry.
Their competitor, with a much smaller site and weaker SEO metrics, was being referenced constantly. I started tracking mentions across different AI systems and realized something was broken in our understanding.
So I did what any good consultant does - I started experimenting. For three months, I tracked every piece of content my clients published, monitoring which pieces got picked up by ChatGPT and which ones were ignored.
The first surprise came from my B2B SaaS client's blog content. Their most comprehensive, pillar-page style articles - the ones that ranked #1 on Google - rarely got mentioned by AI. But shorter, more focused pieces that barely cracked the first page were being cited regularly.
Then I noticed something with another client's product documentation. Their technical API docs were appearing in ChatGPT responses more often than their marketing content, despite having zero traditional SEO optimization.
The pattern became clear when I analyzed an ecommerce client's content performance. Their product pages with detailed specifications and clear, factual information were getting AI mentions. But their SEO-optimized category pages with keyword-stuffed content were nowhere to be found.
What really opened my eyes was discovering that content structure mattered more than content length. A client's simple troubleshooting guide was being referenced constantly, while their comprehensive "ultimate guide" was ignored.
This led me down a rabbit hole of testing different content approaches specifically for AI discoverability. I started treating each section of content as an independent "chunk" that could be extracted and synthesized, rather than optimizing entire pages for specific keywords.
Here's my playbook
What I ended up doing and the results.
Based on my experiments, here's the systematic approach I developed for optimizing content for ChatGPT mentions:
Step 1: Chunk-Level Content Architecture
Instead of thinking in terms of pages, I restructure content into self-contained sections. Each section answers a specific question completely, with clear context that doesn't rely on surrounding content. This mirrors how LLMs process and synthesize information.
For my SaaS client, I broke down their comprehensive product guides into modular sections. Each section could stand alone as a complete answer, with proper context and factual accuracy. The result? Their mention rate increased by 300% within two months.
Step 2: Factual Density Over Keyword Density
Traditional SEO focuses on keyword optimization. For AI systems, I optimize for factual density - ensuring each paragraph contains verifiable, specific information rather than fluff or keyword repetition.
I implemented this with an ecommerce client's product descriptions. Instead of keyword-rich marketing copy, we focused on technical specifications, use cases, and factual comparisons. ChatGPT started citing these pages when users asked about product recommendations.
Step 3: Citation-Worthy Information Hierarchy
AI systems prefer content that reads like reference material. I structure information with clear hierarchies: definitions, explanations, examples, and specific data points. This makes it easy for AI to extract and attribute information accurately.
Step 4: Multi-Modal Content Signals
While AI can't "see" images directly, I've found that content with proper alt text, structured data, and table formats performs better. The structured approach to presenting information seems to signal authority to AI systems.
For one client, I converted their text-heavy comparison content into properly formatted tables with clear headers. Not only did this improve user experience, but ChatGPT started referencing these comparisons regularly.
Step 5: Semantic Completeness
Rather than targeting specific keywords, I ensure content covers topics semantically complete. This means addressing related concepts, definitions, and context that AI might need to provide comprehensive answers.
Content Structure
Each section must function as a standalone answer with complete context - no dependency on surrounding content for comprehension.
Factual Density
Replace keyword optimization with factual density - every paragraph should contain specific, verifiable information rather than marketing fluff.
Reference Format
Structure content like reference material with clear hierarchies, definitions, examples, and data points for easy AI extraction.
Semantic Coverage
Cover topics semantically complete with related concepts, definitions, and context rather than targeting isolated keywords.
The results from this approach have been consistently surprising across different clients and industries.
For my B2B SaaS client, we tracked a 300% increase in ChatGPT mentions within two months of implementing the chunk-level restructuring. More importantly, these mentions were accurate and contextually relevant, not generic references.
The ecommerce client saw their products being recommended by ChatGPT when users asked for specific use cases, despite having lower Google rankings than competitors. This translated to a measurable increase in "AI-driven" traffic - users who mentioned finding them through ChatGPT.
But the most telling result came from tracking mention quality. Traditional SEO-optimized content that did get picked up by AI was often quoted out of context or with inaccuracies. The reference-style content I optimized was being cited more accurately and completely.
What surprised me most was the speed of results. Unlike traditional SEO, which can take months to show impact, AI optimization seemed to work much faster. Content structured properly was getting picked up within weeks, not months.
The approach also had an unexpected benefit: it improved traditional SEO performance. Google's algorithm increasingly favors content that directly answers user questions, so the reference-style formatting boosted search rankings as well.
What I've learned and the mistakes I've made.
Sharing so you don't make them.
Here are the key insights from six months of AI optimization experiments:
Quality beats quantity every time. One well-structured section outperforms ten keyword-stuffed paragraphs. AI systems seem to prioritize accuracy and completeness over content volume.
Think like a reference librarian, not a marketer. Content that reads like an encyclopedia entry performs better than content that reads like a sales page. AI wants to cite authoritative information, not promotional material.
Context is everything. Each section needs to provide enough context to be understood independently. AI doesn't read your entire page - it extracts relevant chunks.
Structure signals authority. Proper formatting, clear hierarchies, and logical information flow seem to signal credibility to AI systems, even without traditional authority signals like backlinks.
Semantic completeness matters more than keyword targeting. Covering a topic thoroughly from multiple angles outperforms targeting specific keyword phrases.
Speed of optimization. AI optimization works faster than traditional SEO but requires more fundamental content restructuring rather than surface-level tweaks.
Accuracy is paramount. AI systems seem to favor content with factual accuracy over content with SEO optimization. One inaccurate statement can hurt your chances of being cited.
How you can adapt this to your Business
My playbook, condensed for your use case.
For your SaaS / Startup
For SaaS companies looking to optimize for AI mentions:
Restructure documentation into standalone, complete answers
Focus on use case coverage rather than feature promotion
Create comparison content that positions your solution contextually
For your Ecommerce store
For ecommerce stores optimizing for AI discoverability:
Emphasize product specifications and technical details over marketing copy
Structure comparison data in tables and clear hierarchies
Cover use cases comprehensively with specific, factual information