Your Wix Site Is Missing the Two Files That Control AI Search Visibility

Three controls look like AI visibility infrastructure. None of them work the way Wix implies. The platform does not tell you this. The setup screens show nothing.

Check Your Own Site First

Before reading further, open a browser tab and type yourdomain.com/robots.txt. If you see a list of rules, Wix has one. Then type yourdomain.com/llms.txt. Most Wix service sites return a blank page. That blank page is the problem.

Illusion 1: Your robots.txt Is Blocking AI Crawlers

Wix published a guide on blocking AI crawlers. It names four bots: CCBot, GPTBot, ChatGPT-User, and BingAI. That list is already outdated. It misses ClaudeBot, Claude-SearchBot, Claude-User, PerplexityBot, OAI-SearchBot, and Google-Extended. If you followed that guide, you blocked a fraction of the crawl traffic hitting your site.

The Bots Wix Names and the Ones It Forgets

Anthropic runs three separate bots with three separate jobs. Here is what each one does and what the right move is for a typical service firm:

Most robots.txt guides treat training bots and search bots as the same thing. They are not. Blocking ClaudeBot while allowing Claude-SearchBot keeps you out of Anthropic's training data and inside Claude's search answers. Wix's guide makes no distinction between the two. The how to rank in AI search guide covers per-crawler rule structure in detail.

Two Modes: Standard Crawlers vs. Live Retrieval

There are two distinct operating modes for AI systems.

Standard crawlers like GPTBot and ClaudeBot index your site on a schedule. They do honor robots.txt. If you block them there, they stay out.

Live retrieval tool calls are different. When a user asks ChatGPT, Claude, or Perplexity a question in real time, those systems can call a web_fetch tool that retrieves your page directly. Google's own documentation confirms that robots.txt does not apply to those direct requests. The file was never a wall for real-time retrieval. It is an honor system for scheduled indexing.

Not every crawler even honors that system. On August 4, 2025, Cloudflare documented Perplexity operating a stealth crawler that impersonated Chrome on macOS. It rotated IPs and ASNs specifically to bypass robots.txt. Cloudflare de-listed Perplexity as a verified bot entirely.

Blocking the Wrong Bot Removes You from AI Answers

If you block OAI-SearchBot instead of GPTBot, you remove your site from ChatGPT Search while leaving your content in OpenAI's training pool. That is the opposite of what most founders want. Run a crawler configuration audit before you assume your current setup is doing what you think.

Illusion 2: Wix Auto-Generates Your llms.txt

robots.txt and llms.txt do completely different jobs. robots.txt operates at crawl time. It controls how bots discover and index your site. llms.txt operates at inference time. It guides what an AI cites while writing an answer to a user's question right now. Most Wix founders have never heard of llms.txt. Many assume the platform handles it. Neither assumption is safe.

Who Actually Gets One

Wix auto-generates llms.txt for exactly one group: Premium eCommerce sites in the US, in English. Every other site tier gets nothing. No file. No placeholder. If you run a service business, a local practice, a B2B firm, or anything outside that narrow slice, Wix has not created this file for you.

Jeremy Howard published the llms.txt standard on September 3, 2024. The format uses Markdown, an H1 title, an optional blockquote summary, and H2-delimited sections of URL links with descriptions. Google has already indexed tens of thousands of these files globally and added an llms.txt check to Chrome Lighthouse. This is infrastructure now, not an experiment.

The Edit Trap

If you do have a Wix-generated llms.txt and you manually edit it, Wix permanently disables auto-updates. You end up with either a stale custom file or no version control at all. There is no middle ground.

The File Wix Never Creates

llms-full.txt is the companion file. It includes the full text of your key pages so AI systems can ingest your content without crawling individual URLs. Wix does not generate llms-full.txt for any site tier.

Without both files in place, AI systems answering questions in your category have no structured source to pull from. Your competitors on managed stacks do.

Workaround for Non-Premium Wix Sites

If Wix does not generate your llms.txt automatically, you can serve a custom file using a Cloudflare Worker. The Worker intercepts requests to yourdomain.com/llms.txt and returns a file you control. This is not a native Wix feature. It requires a Cloudflare account and a domain routed through Cloudflare DNS. It does work, and it gives you the root-path placement that AI crawlers expect.

Illusion 3: Wix Handles Your Structured Data

Wix does generate structured data automatically, but only for five content types: blog posts, Wix Stores product pages, Wix Bookings services, Wix Events pages, and homepages with Local Business markup. That is it.

Everything else requires you to add JSON-LD manually. JSON-LD is a block of code, injected into your page's <head>, that tells AI and search systems exactly what your page is about. On Wix, you add it through Settings, Advanced, then Custom Code.

What Wix Skips

Services pages. Custom landing pages. About pages. FAQ sections. All of these get zero native schema on Wix. You have to build it yourself, paste it in, and maintain it as your content changes.

Most firm owners never do.

The Caps Nobody Warns You About

Even when you add structured data on Wix, you hit two hard limits: each markup entry caps at 7,000 characters, and you can only add five markups per page total.

For a professional services firm with a detailed offerings page, those are not theoretical limits. A thorough ServiceSchema with multiple service types and delivery details fills 7,000 characters fast. Once it is full, you start cutting.

Why Your Services Page Hurts the Most

Your services page is your lead generation engine. It also gets zero automatic structured data from Wix.

FAQPage schema is one of the highest-impact markup types for AI citations. Wix forces you to implement it manually on any page outside its native five content types. Most founders never get around to it. That gap compounds every week.

What the Full Stack Actually Looks Like

We audit Wix sites for professional services clients. The same gaps appear every time. Here is what properly implemented AI visibility actually includes:

• Per-crawler robots.txt rules that block training bots without cutting off search indexing
• A properly formatted llms.txt and llms-full.txt at the site root
• Organization schema with sameAs links across every major directory
• FAQPage schema on every relevant page
• Canonical URLs set correctly across the entire site
• Log file analysis showing which AI agents hit your site and what they read

The last item is the one no platform gives you. Wix launched its AI Visibility Overview tool on July 16, 2025. Watching citation counts fall is not the same as knowing why they fall. Log file analysis tells you whether GPTBot is actually crawling your service pages, which sections it reads, and where it stops. No Wix dashboard surfaces that.

A log gap is also where you find crawl problems that look invisible everywhere else. A bot could be hitting your homepage and ignoring every page below it. You would never know without the raw access logs.

Research from Search Engine Land shows 40-60% of sources cited by major AI platforms rotate every month. A source that ranks in AI answers in May is frequently gone by June. That is why one-time setup is not enough. Professional AI visibility consulting is a maintenance cycle. The firms winning AI-referred traffic treat it that way.

Frequently Asked Questions