Your Website Has Two Audiences. Most Businesses Only Write for One.

Every small business owner I talk to has thought about their website from the customer’s perspective. They have written the copy, chosen the photos, mapped the navigation. They know what a visitor sees when they land.

Almost none have thought about it from Google’s perspective. Or from ChatGPT’s.

The gap between those two perspectives is where most search visibility is lost.

Your site has two audiences, and they read differently

When a customer visits your site, they are reading copy. They are evaluating trust. They are deciding whether to call.

When Google crawls your site, it is not reading copy. It is parsing signals. Specifically, it is trying to answer three questions: what does this business do, where does it operate, and is it real and established enough to surface with confidence?

AI engines like ChatGPT and Perplexity work differently from traditional search, but they are asking the same underlying questions. A Hacker News thread from February 2026 put it clearly: these systems do not rank pages the way search engines do. They select sources, summarize them, and recommend them directly. The selection criteria are authority, structured data, and how cleanly the content answers a specific question.

Both audiences, traditional search and AI, are looking for signals your content does not naturally produce on its own. Those signals have to be built deliberately.

The signal layer most sites are missing

When I run the six-category diagnostic on a small business site, the pattern I see most consistently is this: the site looks good, the copy is solid, and the machine-readable layer is either absent or so minimal it communicates almost nothing.

Schema markup is the primary mechanism through which a site tells search engines and AI crawlers what it is in a language they can parse without interpretation. A LocalBusiness schema block, for example, tells Google the business name, address, phone number, service area, hours, and the categories it operates in. It tells AI crawlers the same things, plus it establishes an entity, a named, structured record of a real business with defined attributes.

Without that entity definition, the site is a collection of text. With it, the site becomes a node in a knowledge graph that search engines and AI systems can reference with confidence.

Most small business sites have no schema markup at all. Some have a bare minimum, a name and address. Very few have the full entity definition that tells AI engines everything they need to recommend the business for a category query.

What “entity confidence” means in practice

The electrician I wrote about in an earlier essay had this problem in its clearest form. Ask ChatGPT about his business by name and you get a complete, accurate answer. The branded entity exists. Ask ChatGPT who the best electricians in Vancouver are, and he does not appear. The category entity does not exist.

That is the distinction between branded AEO and category AEO. Branded: AI knows who you are when asked. Category: AI knows what you do and recommends you when someone asks for that service.

Category AEO requires a richer signal set. It requires service-area schema that clearly defines where the business operates. It requires service-type markup that names the specific things the business does. It requires content depth on those service types: actual pages answering the questions customers type, not just a single home page that mentions the service in passing.

AI systems use a multi-stage process to decide what to cite: retrieve candidate sources, evaluate each for authority and structured data, filter for fact-checkability. A business with no schema and no service-area signals does not make it through the filter stage. It is not that the AI found the business and decided not to recommend it. The business was never in the candidate set.

The signals that actually move the dial

Three in priority order, from what I see on audits.

LocalBusiness schema with service area. This is the foundational entity signal. It names the business, defines the geography, lists the service categories, and gives AI engines a structured record to reference. Without this, a site has no machine-readable identity. It is structurally invisible to AI citation regardless of how good the copy is.

Service-specific content. One home page that mentions “electrical services” is not content depth. It is a mention. Topical authority comes from dedicated pages or articles that answer the specific questions customers type: what does an EV charger installation cost in Vancouver, how long does a panel upgrade take, what is the difference between 100-amp and 200-amp service. Competitors who have ranked for years have thirty to a hundred pieces of this content. It is built over time and there is no shortcut, but starting now is always earlier than starting later.

Crawl access for AI bots. GPTBot, ClaudeBot, and PerplexityBot are the AI crawlers that index content for ChatGPT, Claude, and Perplexity respectively. By default, a site that blocks crawlers in robots.txt, or that was built with a blanket disallow rule, is invisible to all of them. I check this on every audit. It comes up more than it should.

The thing nobody tells you in the design brief

When a site gets built, the brief covers copy, design, and functionality. It almost never covers the machine-readable layer. That layer is not visible to visitors, it does not affect how the site looks, and it does not come up in the conversation between a business owner and a designer.

So it gets skipped. Not because anyone decided to skip it. Because nobody knew to ask for it.

The result is a site that works perfectly for the one audience it was built for, and does almost nothing for the other. In 2026, that second audience, search engines and AI systems, is the one that determines whether new customers find the business at all.

---

Ian Chang is the founder of SyncSpark, a Vancouver agency specializing in SEO, AEO, and AI search optimization. Run the free Website Scorecard to see how your site scores across five signal categories: speed, mobile, search foundation, AI visibility, and Google rankings.