How AI Search Engines Reach Your Customers (And Why Your Website Is Not Built for It)

Millions of people now use ChatGPT, Gemini and Perplexity as their primary search engine. Instead of a list of ten blue links, they get a direct answer to their question. "Which restaurant in Amsterdam has the best steak?" "What is the best webshop for tiles in the Netherlands?" "Which contractor in London is reliable for a bathroom renovation?" AI gives one answer, sometimes two or three. Not ten options to sift through yourself.

That changes everything for businesses that depend on online discoverability. With Google, you could afford to be in position 5. You got fewer clicks than position 1, but you were visible. With AI search engines, you either exist or you do not. There is no page 2. There are no ads you can buy your way to the top with (not yet, at least). The AI chooses who gets recommended and who gets ignored. If you are not in that answer, you do not exist for that searcher.

The shift is already measurable. Organic clicks from Google have dropped by 61% for queries where an AI overview is shown. That traffic is not vanishing into thin air. It is moving to AI platforms that answer directly, without the user ever visiting a website. Unless the AI names your website as a source. Then you do get the visit, and it is often higher quality: the user arrives with a specific intent, not to compare.

This is not a trend that becomes relevant in five years. This is happening now. Every month, the share of people using AI as a search engine grows. And most websites are simply not built for it.

Every day you wait, a competitor gains ground you will not get back. The first businesses to adapt to AI search are building a lead that compounds over time. In six months, the gap will be too wide to close.

AI search engines work fundamentally differently from Google. Google ranks pages based on links, keywords, domain authority and hundreds of other signals. The result is a ranked list. AI systems do something different: they read your website, try to understand the content, and decide whether you are the best answer to a specific question a user asks.

That "understand" is the crucial word. An AI system does not read like a human. It looks for structured information it can interpret and verify. Who is this business? What exactly do they do? Where are they located? What are their products or services? What are the prices? What do customers say about them? What makes them different from competitors? If that information is scattered, vague or poorly structured on your website, the AI cannot do much with it. It lacks the context to select you as a relevant answer.

AI systems favour sources that provide three things:

1. Clarity. Clear, factual information about who you are and what you do. No marketing jargon, no vague promises like "we deliver quality," but concrete facts. "Tile webshop with 5,000+ products, delivery within the Netherlands in 2-3 business days, prices from €18.95 per m2." That is information an AI can work with.
2. Structure. Information organised in a machine-readable way. Not just aesthetically pleasing for human eyes, but also logical for an algorithm trying to understand what is on your page.
3. Trustworthiness. Consistent information across multiple sources. If your Google Business Profile shows different opening hours than your website, or if your address on your Facebook page differs from your website, you lose credibility with the AI. Inconsistency is a signal of unreliability.

Structured data is a way of organising information on your website so that machines can understand it directly without having to guess. Imagine handing a business card to someone who does not speak your language. The text is there, but the person does not know what your name is, what your phone number is, what your role is and what your company does. Structured data is like a universal format for that business card: every piece of information has a label that works in every language.

In technical terms, this is called "schema markup." It is code that is invisible to visitors on your website, but visible to Google, ChatGPT, Gemini, Perplexity and all other systems that read your site and try to understand it. It explicitly tells them: this is a business, this is the address, these are the services, these are the reviews, this is the price of this product, this is the delivery time.

For a webshop, this means: every product gets structured data with the name, price, stock status, image, rating, description, material, dimensions and availability. When someone asks ChatGPT "where can I buy black floor tiles 60x60 in the Netherlands?", the AI needs that structured information to give your webshop as the answer. Without that data, the AI has to guess based on loose text on your page, and that almost always goes in favour of a competitor who has set it up properly.

For a service business, it means: your company name, location, services, service area, opening hours, reviews and contact details are all available in structured form. So that when someone asks "which web agency in the Netherlands builds custom webshops?", the AI can find and mention you.

Without structured data, your website is a book without a table of contents. The information is there, but nobody finds it at the right moment. And in the world of AI search, "not being found" is the same as "not existing."

Wix, Squarespace and standard Shopify themes offer minimal support for structured data. They add the absolute basics (page title, meta description, maybe a basic product schema), but miss the depth that AI search engines need to truly understand your website and recommend it.

A standard Shopify product page tells Google and AI systems: "this is a product with a name and a price." That is it. A well-optimised custom page tells them: "this is a floor tile by brand X, material porcelain, size 60x60cm, colour anthracite, surface finish matte, slip resistance class R10, frost-resistant yes, suitable for underfloor heating yes, price €34.95 per square metre, 247 units in stock, rating 4.8 out of 5 stars based on 89 reviews, deliverable within 2-3 business days across the Netherlands, return policy 30 days."

The difference is enormous. The first version gives the AI almost nothing to work with. It is like asking a librarian to find a book and saying "it has a blue cover." The second version gives the AI everything it needs to recommend you as the best, most relevant answer. It is like giving the title, the author, the ISBN and the shelf location.

This is not a matter of installing a plugin and being done. Most SEO plugins for template platforms add basic schema markup, but miss product-specific properties, FAQ markup, breadcrumb structure, business information markup, review aggregation and the dozens of other schema types that AI systems use to understand your website. That depth requires custom work. It requires someone who understands how schema markup works and how your business works, so that the right information is structured in the right way.

GEO stands for Generative Engine Optimization, and it is the successor to traditional SEO. Where SEO is about ranking higher in a list of search results, GEO is about being chosen as the answer that AI presents directly to users. The goal is no longer "be at the top of the list." The goal is "be the answer."

The principles of GEO partially overlap with good SEO, but go further on crucial points:

Write answers, not marketing copy. AI systems look for factual, direct answers to specific questions. "Our tiles are the best choice for your project" yields nothing for an AI system. "Our collection includes 340 porcelain floor tiles in sizes from 30x30 to 120x120cm, with prices from €18.95 per square metre, deliverable across the Netherlands within 2-3 business days" gives the AI concrete information to work with. Facts, figures, specifications.

Structure your content around questions people actually ask. People ask AI literal questions in natural language. "What does a bathroom renovation cost?" "Which floor tile is best for a humid space?" "How long does it take to get a webshop built?" If your website answers those questions in recognisable formats (FAQ sections, clear headings, direct answers in the first sentence of each section), you increase the chance that AI cites you as a source.

Build authority through consistency across all platforms. AI systems check information against multiple sources. If your website, your Google Business Profile, your LinkedIn, your industry listings and your social media all show the same information (same name, same address, same services, same opening hours), your trustworthiness as a source rises.

Use schema markup for everything you can structure. Not just products, but also FAQs, reviews, opening hours, location, services, articles, events, tutorials. The more structured data your website contains, the better AI systems understand what you have to offer and when you are the relevant answer.

Imagine: a potential customer asks ChatGPT: "Where can I buy affordable bathroom tiles in the Netherlands with fast delivery?" The AI searches the internet, analyses dozens of websites, and gives an answer. Your webshop is mentioned, or not. What determines that difference?

If your webshop has structured data for every product (name, price, category, material, stock, reviews, delivery time), if your product pages start with a direct, factual description that puts the key information first, if you have FAQs that concretely answer common customer questions, and if your business information is consistent across all platforms (website, Google, social media, industry listings), then you have a serious chance of being mentioned.

If your webshop runs on a standard Shopify theme with generic product descriptions like "beautiful tile, high quality, various colours available" and no schema markup beyond the absolute basics, then you are invisible to those AI systems. Your website technically exists, but the AI does not have enough structured information to conclude that you are the relevant answer.

The same applies to service businesses. When someone asks "which web agency in the Netherlands builds custom webshops?", you only get mentioned if the AI can establish that you are a web agency, that you are in the Netherlands, and that you build custom webshops. Those facts need to be structured on your website, not buried in a paragraph of marketing copy.

A concrete example: a business owner in London asks ChatGPT "which webshop has the best prices for ceramic floor tiles with fast delivery?" The AI searches the internet and gives two or three answers. If your webshop has structured data with product prices, delivery times and stock status, you have a chance of being recommended. Without that data, you simply do not exist in that conversation.

This is not future talk. This is happening now, every day, millions of times. Perplexity processes millions of search queries per day. ChatGPT has hundreds of millions of active users. Google itself shows AI overviews for a growing percentage of all search queries. The shift is not coming. The shift is already underway, and every business that does not move with it is losing potential customers every day to competitors that do.

At Nurani, we have been optimising for AI search engines since before anyone in the Dutch market was talking about it. We build every website and webshop with structured data as the foundation, not as an afterthought. That is a deliberate choice we make on every project, because we believe that discoverability in AI search engines will be just as important as a Google ranking within 12 months. It is not something we "add later" with a plugin. It is part of how we build.

Concretely, that means: every page has schema markup that tells AI systems what the content is, who it is intended for, and how it relates to the rest of your website. Every product page is a complete structured answer to potential customer questions. Every FAQ section uses FAQ schema so AI can cite it directly. Every blog post is written with the first sentence as an extractable answer, in exactly the format AI systems prefer to cite.

We build with Next.js, a framework that supports server-side rendering. That means the content of your page is immediately available when a search engine or AI crawler visits your site, without them needing to execute JavaScript to see the content. This is a technical detail that makes a big difference: many template platforms load content via client-side JavaScript, which means AI crawlers sometimes see an empty or incomplete page. With server-side rendering, the full page with all content and all structured data is immediately available.

For every webshop, we implement product schema, FAQ schema, organisation schema, breadcrumb schema, review schema and local business schema. We ensure the information on your website is consistent with your Google Business Profile and other online listings. We write content that answers the questions your customers ask.

Ask your current web agency about GEO. Chances are they have never heard of it. No other web agency in the Netherlands positions GEO as a core part of their build process. Most have never heard of it. We do it because we see how search behaviour is changing, and because we believe our clients should not fall behind on that. If you get a website built today that is not prepared for AI search, you are building something that will already be outdated within a year. Meanwhile, your competitors who are found by ChatGPT and Perplexity are capturing customers every day that you will never even see.

Want to know how visible your website is to AI search engines? That conversation starts with a free strategy session.