Is Your Store Invisible to AI? 5 Signs to Check Right Now

If you’ve never checked whether AI agents can actually read your store, the answer is probably: they can’t — or not well enough to recommend your products.

ChatGPT, Perplexity, Gemini and Google AI Overview are answering millions of shopping questions every day. When a user asks “what’s the best sustainable yoga mat under €80?”, one of two things happens: your store appears in the answer, or a competitor’s does. There is no middle ground.

The stores that appear aren’t necessarily the biggest or oldest. They’re the ones with the right technical signals — signals that most merchants have never heard of, let alone set up.

Here are the five signs your store is invisible to AI, and how to check each one.

Sign 1: AI crawlers are blocked in your robots.txt

Your robots.txt file tells web crawlers what they can and cannot access. Many stores have rules that block all unrecognized bots — which includes GPTBot (ChatGPT’s crawler), ClaudeBot (Anthropic/Claude), and PerplexityBot.

How to check: Visit yourdomain.com/robots.txt in your browser. Look for any of these entries:

User-agent: *
Disallow: /

Or the absence of explicit allow rules for:

User-agent: GPTBot
User-agent: ClaudeBot
User-agent: PerplexityBot

If GPTBot can’t crawl your store, ChatGPT’s knowledge of your catalog is either zero or based on outdated cached data. Same for every other AI system.

The fix: Add explicit allow rules to your robots.txt:

User-agent: GPTBot
Allow: /

User-agent: ClaudeBot
Allow: /

User-agent: PerplexityBot
Allow: /

This alone can move the needle significantly — it’s the highest-impact single change most stores can make.

Sign 2: No structured data on product pages

When a human reads your product page, they understand the context automatically. When an AI agent reads it, it needs explicit machine-readable signals to understand this is a product, it costs X, it’s in stock, it’s made for Y use case.

That’s what JSON-LD structured data provides — and most Shopify stores either have minimal schema or none at all.

How to check: Right-click any product page, click “View Page Source”, and search for application/ld+json. You should see blocks like:

{
  "@type": "Product",
  "name": "...",
  "offers": { "price": "..." }
}

If you find nothing, or only a basic WebPage schema, your products are invisible to structured data parsers.

The impact: AI agents that can’t parse your product schema have to reconstruct the information from your page’s HTML — titles, paragraphs, image filenames. This produces low-confidence outputs that get filtered or hedged in AI answers.

Sign 3: No llms.txt file

llms.txt is a machine-readable index file that tells AI models what your site contains, which pages are most important, and how to interpret your catalog. Think of it as a sitemap.xml — but designed for AI systems rather than search engine crawlers.

How to check: Visit yourdomain.com/llms.txt. If you get a 404 error, it doesn’t exist.

Most stores don’t have one. This means AI models have no structured reference for your catalog when constructing answers — they rely entirely on what they’ve seen in training data or crawled recently, which is incomplete at best.

The fix: You can generate a free llms.txt for any domain using the Krawfly llms.txt generator. It takes under a minute, no signup required. Publish the file at the root of your domain and AI agents can immediately reference it.

Sign 4: Thin or missing meta descriptions

Meta descriptions aren’t just for Google anymore. AI agents use page metadata to quickly understand what a page is about before deciding whether to read the full content. A missing or auto-generated meta description signals low content quality.

How to check: Search your brand name on Google. Look at the text snippet below each result title. If it reads like a random excerpt from the page, your meta descriptions are either missing or not set manually.

On a product page level: view source and search for <meta name="description". The content attribute should be a clear, specific sentence describing the product — not a generic template.

What AI agents need: A description that answers what is this product, who is it for, and what problem does it solve — in one or two sentences. Vague descriptions like “Premium quality product from our collection” give AI nothing to work with.

Sign 5: Your score tells the full picture

The four signs above catch the most common issues. But there are more: hreflang tags for multilingual stores, canonical URL configuration, image alt text quality, sitemap structure, and the overall semantic context of your product descriptions.

The fastest way to see all of them at once: paste your store or product URL into the free AI Visibility Score tool. You’ll get a 0–100 score broken down across five signal categories, with each missing signal ranked by how many points fixing it would add.

Most stores we scan score between 30 and 55. A score below 40 means AI agents are effectively guessing about your products. Above 75, you’re in a strong position for AI-driven recommendations.

---

What to do if your score is low

Start with the three fixes that cost nothing and have the highest impact:

1. Update robots.txt — allow GPTBot, ClaudeBot, PerplexityBot. Immediate effect.

2. Create llms.txt — use the free generator, publish at your domain root. Done in 2 minutes.

3. Add JSON-LD Product schema — if your platform doesn’t generate it automatically, add it manually to your theme or via a script tag. Krawfly automates this for Shopify stores across the full catalog.

The difference between a store at 35 and one at 75 isn’t catalog size or brand authority — it’s whether someone set up these signals. Most of them take under an hour to fix manually. For large catalogs, Krawfly handles the generation, publishing, and monitoring automatically.

Check your AI Visibility Score — free, no signup →