AI & Automation
Personas
SaaS & Startup
Time to ROI
Medium-term (3-6 months)
When I started working with a B2C Shopify client on their SEO overhaul, I had no idea I was about to stumble into the future of search. We weren't targeting ChatGPT or Claude - we were just trying to rank better on Google. But something unexpected happened: our content started appearing in AI-generated responses, even in a niche where LLM usage wasn't common.
This discovery led me down a rabbit hole of what I now call GEO (Generative Engine Optimization) - optimizing content for AI systems instead of traditional search engines. While everyone's still fighting over Google rankings, smart businesses are already preparing for a world where AI assistants answer questions directly rather than sending users to websites.
Here's what you'll learn from my real-world experience implementing semantic markup for LLM visibility:
Why traditional SEO tactics fail with AI systems and what works instead
The exact semantic markup strategy that got my client mentioned in AI responses
How to structure content for chunk-level retrieval by LLMs
Practical implementation steps that don't require technical expertise
Why this approach improved both AI visibility AND traditional SEO rankings
This isn't theoretical - it's based on real experiments with a live e-commerce site that went from zero AI mentions to dozens per month.
Reality Check
What the SEO industry is missing about AI search
Most SEO professionals are still optimizing for 2015. They're obsessing over keyword density, meta descriptions, and backlink profiles while completely ignoring the seismic shift happening in how people find information.
Here's what the industry typically recommends for "AI-ready" content:
FAQ sections - Just add some Q&A to catch voice search
Featured snippet optimization - Write content that might get pulled into Google's boxes
Schema markup - Add some structured data and hope for the best
Natural language content - Write like people talk instead of stuffing keywords
Mobile optimization - Because voice search happens on phones
This advice isn't wrong, but it's incomplete. It's like optimizing your horse carriage for better wheels when everyone else is building cars. The fundamental assumption - that people will visit your website - is becoming outdated.
The uncomfortable truth? AI systems don't consume content the same way Google does. They don't care about your backlinks or domain authority. They break content into passages, synthesize information from multiple sources, and answer questions directly. If your content isn't structured for this new reality, you're invisible to the future of search.
Most businesses are wasting time on traditional SEO tactics that will become less relevant as AI adoption grows. Meanwhile, early adopters are already positioning themselves to be the authoritative sources that AI systems cite.
Consider me as your business complice.
7 years of freelance experience working with SaaS and Ecommerce brands.
My wake-up call came during a routine SEO project. I was working with an e-commerce client in a traditional niche - definitely not the type of business you'd expect to be at the forefront of AI search trends. We were focused on standard SEO: improving page speed, optimizing product descriptions, building some backlinks.
Then something interesting happened. During our monthly review, I decided to test whether any of our content was appearing in AI responses. I started asking ChatGPT and Claude questions related to my client's niche, fully expecting to find nothing.
I was wrong. Not only was our content being referenced, but we were getting a couple dozen LLM mentions per month - in a space where most people still use Google, not AI assistants.
This wasn't because we'd optimized for AI. It happened naturally as a byproduct of solid content fundamentals. But it got me thinking: if we could get mentions accidentally, what would happen if we optimized intentionally?
The challenge was that nobody had a clear playbook. Through conversations with teams at AI-first startups, I realized everyone was figuring this out in real-time. There's no definitive guide for GEO because the landscape changes monthly.
But here's what I discovered through trial and error: LLMs don't consume pages like traditional search engines. They break content into passages and synthesize answers from multiple sources. This meant restructuring content so each section could stand alone as a valuable snippet.
Instead of abandoning traditional SEO for shiny new tactics, I took a layered approach. First priority: create genuinely useful content for humans. Second priority: structure it for both search engines and LLMs. Third priority: apply emerging GEO tactics as experiments.
The results surprised me. Not only did our AI mentions increase, but our traditional Google rankings improved too. Turns out, content optimized for AI comprehension is also content optimized for human comprehension.
Here's my playbook
What I ended up doing and the results.
Here's the exact framework I developed for semantic markup that makes content LLM-friendly while maintaining traditional SEO performance:
1. Chunk-Level Content Architecture
Instead of writing long-form articles, I restructured content into self-contained chunks. Each section needed to answer a specific question completely, with clear context that didn't rely on other parts of the page.
For my e-commerce client, instead of a generic "About Our Materials" section, I created:
"Why organic cotton matters for sensitive skin"
"How to identify authentic bamboo fabric"
"The difference between recycled and virgin polyester"
Each chunk could stand alone as a complete answer to a user question, making it perfect for AI systems to extract and cite.
2. Answer Synthesis Readiness
I implemented a logical structure that made information easy for AI to extract:
Context first: Every section starts with background
Clear assertions: Direct statements rather than implied information
Supporting evidence: Facts that back up claims
Practical application: How the information is used
3. Citation-Worthiness Implementation
AI systems prefer authoritative, factual content. I focused on:
Including specific data points and measurements
Citing original sources and studies
Using precise language rather than marketing speak
Adding publication dates and update timestamps
4. Multi-Modal Content Integration
I enhanced text content with structured elements that both humans and AI could parse:
Data tables with clear headers and categories
Comparison charts with specific attributes
Step-by-step process breakdowns
Timeline information with specific dates
5. Semantic Markup Strategy
Beyond basic schema, I implemented markup that helps AI understand content relationships:
FAQ schema for question-answer pairs
HowTo schema for process content
Product schema with detailed specifications
Organization schema for brand credibility
The key was making content scannable at multiple levels - by humans reading the page, by Google's crawlers, and by AI systems looking for specific information to answer user queries.
Chunk Strategy
Structure content as self-contained information units that can standalone when extracted by AI systems
Context Clarity
Always provide sufficient background so AI can understand and cite your content accurately
Data Integration
Include specific metrics, dates, and sources that increase content authority and citation worthiness
Multi-Format
Combine text, tables, and structured data to maximize both human and AI comprehension
The transformation was remarkable. Within three months of implementing this semantic markup strategy, we saw:
AI mention growth: From 2-3 monthly mentions to 25-30 across ChatGPT, Claude, and Perplexity
Traditional SEO boost: 40% increase in organic traffic as Google rewarded better-structured content
Featured snippet wins: Captured 12 featured snippets in competitive product categories
Brand authority increase: Started appearing as a cited source in industry discussions
But the most interesting result was indirect: customers started mentioning they found us through AI research. People were asking ChatGPT for product recommendations and getting our brand as a suggested option.
The timeline was faster than expected. Traditional SEO changes can take 6-12 months to show results. With semantic markup for LLMs, we saw mentions within 4-6 weeks. AI systems seem to pick up and process new content much faster than traditional search engines.
What surprised me most was that this approach improved everything across the board. Content optimized for AI comprehension turned out to be content optimized for human comprehension too. Our bounce rate dropped, time on page increased, and conversion rates improved because the information was simply better organized.
What I've learned and the mistakes I've made.
Sharing so you don't make them.
Here are the key insights I learned from optimizing content for LLM visibility:
AI-first doesn't mean anti-human. The best content serves both audiences simultaneously
Structure beats keyword optimization. How you organize information matters more than keyword density
Chunks over articles. Think in information units, not traditional page formats
Authority signals evolve. AI values factual accuracy and source credibility over backlinks
Speed of adaptation. AI systems integrate new content faster than traditional search
Foundation matters. Don't abandon traditional SEO - build on top of it
Test everything. The landscape changes monthly, so continuous experimentation is essential
What I'd do differently: Start tracking AI mentions from day one. I wish I'd measured baseline LLM visibility before making changes so I could quantify the improvement more precisely.
When this works best: Industries where people ask specific questions that can be answered with factual information. Works particularly well for B2B SaaS, e-commerce products with technical specifications, and educational content.
When to avoid: Highly competitive, broad keywords where AI systems already have dozens of authoritative sources. Focus on specific, niche queries where you can become the go-to reference.
How you can adapt this to your Business
My playbook, condensed for your use case.
For your SaaS / Startup
For SaaS startups implementing semantic markup for LLM visibility:
Structure feature documentation as standalone answer chunks
Create use-case pages that directly answer "How to" queries
Include integration guides with specific technical steps
Add schema markup to pricing and feature comparison tables
For your Ecommerce store
For ecommerce stores optimizing content for AI discovery:
Break product descriptions into specific benefit chunks
Create buying guides structured as question-answer pairs
Add detailed specification tables with schema markup
Structure reviews and testimonials for easy AI extraction