How AI-ready are small business websites in Wassenaar? 2026 study

When someone asks ChatGPT, Gemini or Google’s AI Mode for “a good bakery in Wassenaar”, the answer names two or three businesses. Which ones depends heavily on what AI assistants can read on their websites. So in October 2026 I measured 63 websites of businesses in Wassenaar.

Key findings

  • No website meets all six basics of AI visibility. Of the 42 sites where I could measure everything, half meet three or fewer.
  • 31% of reachable sites refuse one or more AI crawlers at the server (13 of 42), usually with a “forbidden” error (403). None of those 13 asks for that in its robots.txt.
  • Only 16% tell machines what kind of business they are with structured data for a local business, such as LocalBusiness or Restaurant.
  • 22% have an FAQ with customer questions; exactly one site also marks it up for machines.
  • The good news: 97% of sites show their text without JavaScript, and 73% list a phone number and address. The foundation is often there, just not finished.

What did I measure?

Six basics that decide whether an AI assistant can find, understand and mention your business:

Basic Wassenaar websites that have it
Readable for AI crawlers (not blocked, text without JavaScript)* 62%
Phone and address on the website 73%
Opening hours on the website 68%
Loads within 4 seconds 81%
FAQ with customer questions 22%
Structured data for a local business 16%

* Measured on the 42 sites that were reachable during the crawler test; the other basics on all 63.

In addition: 46% mention “Wassenaar” in the homepage title, 52% have a meta description and 10% have an llms.txt file.

One in three sites keeps AI crawlers out, often without knowing

This was the biggest surprise. I requested every homepage as a normal browser and then as four AI crawlers: OpenAI’s GPTBot and OAI-SearchBot, Anthropic’s ClaudeBot and PerplexityBot. Of the 42 sites that served a normal visitor properly, 13 refused at least one AI crawler. On 11 sites the pattern was identical: GPTBot and ClaudeBot get “403 forbidden”, the other two are let in.

None of those sites asks for this in its robots.txt, the file where you state which bots are welcome. The pattern appears on sites built with different systems, such as WordPress and Joomla. That points to a default setting of a firewall or hosting provider, not a deliberate choice by the business owner.

Does it matter? Partly. GPTBot mainly collects training data; for ChatGPT search OpenAI uses OAI-SearchBot, which got in almost everywhere. But if you want AI assistants to learn what your business does, you are now shutting them out without noticing. Above all, it is something you should decide yourself.

Machines often don’t know what kind of business you are

Just over half of the sites (56%) have structured data (schema.org), but mostly only generic data that a website builder adds automatically. Only 16% state in that code what kind of business they are: a restaurant, dentist, estate agent or shop, with address and opening hours. That is exactly the data Google and AI assistants use to match you to a local question.

Customer questions rarely get answered

AI assistants answer questions. A site with an FAQ gives them ready-made answers: “Can I park?”, “Do I need to book?”, “What does a treatment cost?”. Only 22% of sites have such a section, and only one marks it up as an FAQ for machines.

Differences by sector

Sector Websites Phone and address Opening hours FAQ Local schema data
Business services 9 100% 78% 44% 22%
Hospitality 15 60% 73% 20% 27%
Health and personal care 14 64% 64% 29% 7%
Shops 22 73% 59% 14% 14%

Business services all list their contact details and most often have an FAQ. Hospitality most often uses structured data, but least often lists both phone and address. In health care only 7% use structured data for a local business. Shops least often have an FAQ or opening hours on the site. Business services here include estate agents, notaries and painters; health and personal care includes dentists, physiotherapists, hairdressers and beauty salons. The groups are small, so treat this as an indication.

What can you do yourself?

  1. Check whether AI crawlers are allowed in. Ask your hosting provider whether their firewall blocks bots such as GPTBot or ClaudeBot, and decide for yourself whether you want that.
  2. Put your address, phone number and opening hours on your website as text, not only in an image or map.
  3. Add structured data for your type of business. In WordPress an SEO plugin can do this; fill in your business type and opening hours there.
  4. Write a short FAQ with the five questions customers ask you most, each answered in the first sentence.
  5. Mention your town in the title and first paragraph of your homepage.

More explanation in Getting found in ChatGPT.

How I did the study

  • Sample: 63 websites of businesses with an address in Wassenaar, spread over more than 30 trades: hospitality, shops, health care, personal care and business services.
  • Website scan: 1 and 2 October 2026, with my own scanner. It reads the homepage and the main pages, both without and with JavaScript.
  • Crawler test: 21 October 2026. One request per crawler per site, with a one-second pause. 21 sites did not serve a page to a normal browser either at that time (maintenance, security or an outage); they are left out of the crawler figures.
  • Privacy: I name no businesses. Only publicly available pages were viewed; nothing was logged into or submitted.
  • Limitations: a snapshot of one town. A crawler refused today may be welcome tomorrow, and a detected FAQ says nothing about the quality of its answers.

Want to know how your website scores? AI visibility is part of my SEO plan, and I’m happy to take a look with you, no strings attached.