I spent the last few weeks building a free tool to answer a single question:
"Can AI actually see your products?"
The results honestly shocked me.
The Test
I built Prodync (prodync.com) — an AI Commerce Visibility platform. And I just shipped a free checker. No signup. No email. Just paste a Shopify product URL and get an AI Visibility Score (0-100) instantly.
Then I started testing. Random Shopify stores. Big brands. Small independents. Stores with solid SEO and professional photos.
What I Found
Most products scored between 30 and 45 out of 100.
Not because the products were bad. But because AI assistants like ChatGPT, Gemini, and Perplexity literally cannot read messy, unstructured product pages.
The most common issues:
- AI crawlers blocked by default in Shopify's
robots.txt - Missing or broken Schema.org structured data
- Product descriptions written for keyword density, not human readability
- Zero attributes beyond price and title
Why This Matters
Some numbers worth sitting with:
- 45% of shoppers now use AI for product discovery
- 62% of those use AI specifically to compare products before buying
- AI-driven retail traffic grew 693% during the last holiday season
- Yet most Shopify stores have no idea they're invisible
What the Checker Does
- Paste any Shopify product URL
- Get your AI Visibility Score (0-100) instantly
- See a breakdown across 5 dimensions: Product Health, Structure, Clarity, Richness, AI Readiness
- Get specific, ranked fixes — like "Add semantic attributes (+18 AI Score)" or "Fix description clarity (+14 AI Score)"
Free. No signup. No BS.
Try It
If you run a Shopify store — or build for one — drop your score in the comments. I genuinely want to know what you find.
Solo founder. Bootstrapped. Building in public.
Top comments (1)
"AI visibility" is a genuinely emerging category and you're early to it - as buyers shift from googling to asking an LLM, whether your product is even retrievable/citable by the model becomes a new SEO, and most stores have done zero to optimize for it (structured data, clear descriptions the model can parse, presence in the sources LLMs actually pull from). "Invisible to AI" is going to be the new "page 2 of Google."
The tool itself is a nice example of the pattern I find most defensible right now: a sharp diagnostic that surfaces a problem people didn't know they had. The natural next step is going from "free checker" to a real product (auth, accounts, billing for the fix-it features) - which is exactly the jump I built Moonshift for: a multi-agent pipeline that takes a prompt to a shipped SaaS on your own GitHub + Vercel, boring-20% as defaults, ~$3 flat per build. First run's free, no card, if the checker earns a paid tier. Cool build - what's your signal for "visible": getting cited in LLM answers, or structured-data presence? Curious how you're measuring invisibility.