The moment I realized SEO was changing for good, I was staring at a Google results page where the first thing a user saw wasn’t a link. It was an answer, synthesized by AI, drawn from sources Google had decided to trust.

I had spent years doing traditional SEO for e-commerce brands: keyword research, on-page optimization, link building, the whole playbook. It worked. But I kept watching search behavior shift. People started asking ChatGPT where to buy things. Perplexity began citing product pages instead of just ranking them. Google AI Overviews started answering “what’s the best X” before a single organic result appeared.

The question wasn’t whether this was happening. It was whether the brands I worked with would be visible in that new layer, or invisible.

Most of them were invisible. Not because their products were bad, but because their sites weren’t structured for machines that synthesize answers. No schema markup telling AI what the product was, who made it, what it cost. No FAQ content written to match the actual questions buyers type into ChatGPT. No structured data letting Perplexity pull a confident citation. They had SEO built for a world that was already shifting underneath them.

So I went deep on what I now call GEO: Generative Engine Optimization. The discipline of making e-commerce content readable, citable, and trustworthy to AI answer engines, not just to crawlers.

How I Work

My approach is hands-on and direct. I log into your store. I scrape your products and catalog. I inject schema markup at the product, category, and site level. I write structured Q&A content that matches the questions real buyers are asking AI tools right now. I build AI-citable pages that give answer engines something concrete to surface when someone asks about your category. Then I hand it back to you: better structured, faster, and discoverable in places your competitors aren’t thinking about yet.

For brands that need more than injection, I rebuild. I work with headless stacks, WooCommerce decoupled with a performant front end, or Medusa paired with Cloudflare, to give stores the page speed and structural clarity that both AI engines and real users reward. My current hands-on proof of concept is gaya.org.il, a 7-year Israeli e-commerce store I am migrating to headless architecture. The before and after on speed and AI citation are documented and public on this site.

Who I Help

I work with e-commerce brands that have real products and real customers but are watching organic traffic flatten and wondering why AI tools never mention them. Usually they are mid-size: enough catalog to benefit from structured data at scale, enough margin to invest in the infrastructure that makes GEO work. Sometimes they need injection into what they already have. Sometimes they need a rebuild. Most need both, in that order.

What I believe, plainly: search isn’t dying, but the front door to discovery is changing. The brands that appear in AI-generated answers will have a structural advantage over the ones that don’t. Getting there requires both content strategy and technical execution. I do both.

If you want to understand where your store stands right now, the gaps are usually findable in an afternoon. That is where I start.