AI & Automation
Personas
SaaS & Startup
Time to ROI
Medium-term (3-6 months)
Here's something nobody talks about: AI can generate 1000 articles in a day, but it can't make a single one discoverable. I learned this the hard way when working with a B2C Shopify client who needed a complete SEO overhaul for their 3000+ product catalog.
The client approached me with what seemed like a straightforward request - revamp their SEO strategy. But here's where it got interesting: they were operating in a traditional e-commerce niche where you wouldn't expect AI content to get traction. Yet somehow, their content was appearing in AI-generated responses despite being in an industry where LLM usage isn't common.
This discovery led me down a rabbit hole that changed everything I thought I knew about content discoverability in the AI era. While everyone's obsessing over gaming ChatGPT or ranking on Perplexity, they're missing the fundamental shift that's already happening.
Here's what you'll learn from my 6-month deep dive into AI content discoverability:
Why traditional SEO thinking breaks in the AI content era
The chunk-level optimization strategy that actually works
How to build content that LLMs naturally reference
The real metrics that matter for AI discoverability
Why quality trumps volume (finally)
If you're still treating AI content like traditional SEO content, you're building beautiful content that nobody - human or AI - will ever find. Let me show you what actually works.
Industry Reality
What every content creator is doing wrong
The content marketing world has lost its mind over AI discoverability. Everyone's trying to "optimize for ChatGPT" or "rank on Perplexity" like these are just new search engines to game. The advice flowing around sounds logical on the surface:
The conventional wisdom everyone's following:
Create FAQ-style content to answer AI queries directly
Stuff keywords into content hoping AI will pick it up
Focus on featured snippet optimization techniques
Build content around "conversational queries"
Optimize for voice search patterns
Here's why this approach is fundamentally flawed: it assumes AI systems work like search engines. They don't. LLMs don't crawl and rank pages - they synthesize information from massive training datasets and live data sources in completely different ways.
The bigger problem? Most businesses are creating more AI content to solve an AI discoverability problem. It's like trying to shout louder in a room that's already too noisy. The result is a sea of generic, AI-generated content that sounds the same and provides no unique value.
Traditional SEO was about matching user intent with content. AI discoverability is about being worthy of citation in a synthesis. That's a completely different game with completely different rules.
While everyone's chasing the shiny new AI optimization tactics, they're missing the fundamental shift: quality and authority matter more than ever, and the platforms that actually drive discovery are the ones nobody's paying attention to.
Consider me as your business complice.
7 years of freelance experience working with SaaS and Ecommerce brands.
My reality check came during a project with a B2C Shopify client. They needed an SEO overhaul for their massive product catalog - we're talking 3000+ products across 8 languages. Standard e-commerce SEO work, nothing fancy.
But something weird happened during my research phase. I was tracking mentions of their brand and products across different platforms when I discovered something unexpected: their content was appearing in AI-generated responses. Not occasionally - we tracked a couple dozen LLM mentions per month.
This was puzzling because they operated in a traditional retail niche where AI usage isn't common. Their customers weren't the type to ask ChatGPT for product recommendations. Yet somehow, AI systems were referencing their content when generating responses about their product category.
The discovery sent me down a research rabbit hole. I started conversations with teams at AI-first startups like Profound and Athena to understand what was actually happening. What I learned changed my entire approach to content strategy.
Everyone was figuring this out in real-time. There was no definitive playbook. The "experts" sharing optimization tactics on LinkedIn were mostly guessing. But the companies actually getting referenced by AI systems? They weren't following any of the conventional wisdom.
I realized we were dealing with a fundamentally different content ecosystem. Traditional SEO metrics became irrelevant. Page authority, backlink profiles, keyword density - none of it mattered for AI discoverability. The content that got picked up by LLMs had completely different characteristics.
This led me to develop what I now call chunk-level optimization - a approach that treats each section of content as a standalone, synthesizable unit rather than optimizing entire pages for search engines.
Here's my playbook
What I ended up doing and the results.
Here's the systematic approach I developed after 6 months of experimentation across multiple client projects:
Step 1: Forget Everything You Know About SEO
AI systems don't consume pages like search engines. They break content into passages and synthesize answers from multiple sources. This meant restructuring content so each section could stand alone as valuable information, complete with context.
Instead of optimizing pages, I started optimizing chunks. Each paragraph needed to be self-contained, factual, and citation-worthy. No more content that only made sense when read sequentially.
Step 2: Build for Synthesis, Not Search
I implemented five key optimizations based on how LLMs actually process information:
Chunk-level retrieval: Each section provides complete context without requiring readers to scroll up for background
Answer synthesis readiness: Information structured in logical, extractable formats
Citation-worthiness: Factual accuracy with clear, attributable sources
Topical breadth and depth: Comprehensive coverage that addresses multiple facets of topics
Multi-modal support: Integration of charts, tables, and visual elements that enhance understanding
Step 3: Quality Over Quantity (Finally)
While competitors were pumping out thousands of AI-generated articles, I focused on creating fewer pieces of genuinely useful content. Each piece needed to serve as a definitive resource on its topic - the kind of content that humans would bookmark and AI systems would naturally reference.
Step 4: Test with Real AI Systems
I started monitoring how content performed across different AI platforms. Not just ChatGPT, but Claude, Perplexity, and even industry-specific AI tools. The goal was understanding which content structures consistently got referenced and why.
Step 5: Build Authority Through Expertise
The content that consistently got picked up by AI systems shared one characteristic: it demonstrated genuine expertise. Not keyword-stuffed expertise, but real insights that couldn't be found elsewhere. This meant going deeper than surface-level advice and sharing actual methods, frameworks, and behind-the-scenes thinking.
For my e-commerce client, this translated into comprehensive product guides that covered not just features, but use cases, comparisons, and technical specifications that weren't available anywhere else. Content that was genuinely useful whether discovered by humans or AI systems.
Key Discovery
AI systems favor content that can stand alone as complete answers, not pages optimized for search rankings.
Chunk Optimization
Each content section must provide full context without requiring additional reading for comprehension.
Authority Signals
Genuine expertise and unique insights consistently outperform keyword-optimized content in AI references.
Testing Framework
Monitor performance across multiple AI platforms to understand which content structures get consistently referenced.
The results spoke for themselves, though they looked completely different from traditional SEO metrics:
For my e-commerce client: We tracked LLM mentions across different AI platforms and saw consistent growth in references. More importantly, the traffic that came from these AI-driven discoveries converted better than traditional search traffic.
The real victory: Content started getting discovered organically by AI systems without any specific "AI optimization" tactics. The focus on genuine value and comprehensive coverage naturally aligned with how LLMs process and synthesize information.
But here's what surprised me most: the principles that made content discoverable to AI also made it more valuable to humans. Better structure, clearer explanations, more comprehensive coverage - these improvements benefited all readers.
Traditional SEO metrics became secondary indicators. Instead of tracking rankings and click-through rates, success meant building content that became a go-to resource in its category.
What I've learned and the mistakes I've made.
Sharing so you don't make them.
After working across multiple client projects and experimenting with different approaches, here are the key lessons that actually matter:
Quality beats optimization every time: AI systems naturally gravitate toward comprehensive, accurate content over keyword-stuffed articles
Structure matters more than length: Well-organized content with clear sections outperforms long-form content that lacks structure
Expertise can't be faked: AI systems are surprisingly good at identifying genuine insights versus regurgitated information
Context is king: Content that provides complete context in each section gets referenced more often
Multi-platform thinking: Don't optimize for one AI system - build content that works across different platforms
Measurement challenges: Traditional analytics don't capture AI-driven discovery - you need new metrics
Sustainable approach: Focus on building lasting value rather than gaming temporary algorithm quirks
The biggest mindset shift: stop thinking about content as something to be found and start thinking about it as something worth citing. That changes everything about how you create and structure information.
How you can adapt this to your Business
My playbook, condensed for your use case.
For your SaaS / Startup
For SaaS companies, focus on:
Create comprehensive use case documentation
Build integration guides that stand alone
Share genuine customer success frameworks
Document your unique methodology clearly
For your Ecommerce store
For e-commerce stores, prioritize:
Detailed product guides with complete specifications
Comparison content that covers multiple options
Use case scenarios for different customer segments
Technical information not available elsewhere