You Optimized for Google. Your Customers Moved to ChatGPT.

Last year, I was waiting to pick up my daughter from ballet class when another parent mentioned she almost did not find the place. She had searched online, found nothing useful, and only heard about it through a friend. The studio owner had sold a large dance school in Canada before relocating to South Florida. She had a professional web presence and a firm managing it. She was doing everything the old playbook recommended.

She was also invisible.

When I checked the studio through ChatGPT, the business did not appear. Several other parents had noticed the same thing. The studio was not surfacing in the searches they actually ran. Not because the product was weak. The teaching was excellent; the fundamentals were being built the right way, and parents who found the place stayed. The problem was not quality. It was discoverability. The two are no longer the same thing.

Search Is Not One Place Anymore

For most of the last decade, digital visibility meant one thing: search engine optimization for Google. Build a well-structured site, earn backlinks, manage your local listing, and you had a reasonable chance of being found. That infrastructure still matters. But it no longer determines who gets discovered.

Customers now find businesses across three distinct surfaces, each operating by different rules.

The first is the traditional search results page. Even there, the rules have changed. In 2024, of every 1,000 Google searches in the US, only 360 resulted in a click to an external website (SparkToro). The rest ended on the results page itself: answers delivered by AI Overviews, featured snippets, or knowledge panels, with no referral out. Google's AI Overviews now trigger on more than 13% of all desktop queries in the US (Semrush, 2025). A business optimized to earn traffic can no longer assume that visibility produces clicks.

The second surface is platform-specific search. Customers do not go to Google for everything. They go to YouTube for tutorials, Reddit for peer reviews, TikTok for recommendations on local services. Research from Adobe found that 25% of Gen Z consumers consider TikTok effective for finding information. Not as a distraction. As a deliberate search behavior. These are not the same people, searching the same way, on the same channel.

The third surface is standalone AI chatbots: ChatGPT, Perplexity, Gemini. According to Orbit Media's 2025 research, 55% of people now use AI tools as their primary or frequent research tool, and 31% prefer AI over traditional search when seeking recommendations. When someone asks ChatGPT for a good ballet studio for beginners in their area, it does not run a keyword search. It synthesizes from what it already knows. A business that has not given the web sufficient, well-structured information about itself does not appear in the synthesis. It simply does not exist in the answer.

Why the Old Playbook Does Not Transfer

The infrastructure built for Google SEO does not automatically make a business legible to AI systems. The criteria are different. Keyword density, backlinks, structured metadata: these were built for a different retrieval model.

A local business visibility study across approximately 350,000 locations found that only 1.2% of businesses were recommended by ChatGPT when asked to suggest a local provider (mshahid.com, 2026). Perplexity recommended 7.4%. Gemini, grounded directly in Google Maps data, reached 11%. Google's traditional local three-pack still surfaced 35.9% of locations. The gap is not marginal. The overlap is partial. Only 45% of businesses visible in traditional search results also appeared in AI recommendations. Being findable in one channel does not mean being findable in the others.

The same study estimated that winning AI visibility is three to thirty times harder than earning a traditional local ranking. That gap exists because AI systems are not ranking pages. They are constructing answers. A business needs to give those systems enough raw material: accurate, consistent, independently corroborated. Enough to be included in a short-list recommendation rather than omitted entirely.

Most small businesses do not know this is how it works. The firms managing their web presence often do not know either. SEO was built for a different era of information retrieval, and the tooling, the auditing, and the advice have not yet caught up to where the searches are actually happening.

What AI Systems Actually Need to Know About You

The information gap is not a mystery. AI search systems look for specific signals to determine whether a business is real, what it does, where it operates, and whether it is worth recommending. Those signals cluster into nine distinct areas.

Identity: Is the business named consistently across every directory, platform, and listing?
Category: Does the internet clearly understand what this business actually is?
Location: Is the geographic footprint explicit and confirmed across multiple sources?
Services: Are offerings described in plain language, not industry shorthand?
Audience: Does available information make clear who this business is for?
Evidence: Are there independent sources (press mentions, community listings, third-party sites) that corroborate this business exists?
Reviews: Do reviews contain descriptive, useful language, or only star ratings with no context?
Technical access: Can AI crawlers read the site? Is the content accessible, or blocked by JavaScript-heavy rendering?
AI query test: When someone asks "Who offers X near me?" or "What is a good X for beginners?" does this business appear?

Applying these questions to the ballet studio exposed the problem quickly. The business name appeared with slight variations across different listings. The services page described class types in language that made sense to dancers, not to parents searching online. There were very few independent mentions anywhere on the web. The reviews that existed were thin: star ratings without the kind of descriptive language ("good for young beginners," "focuses on fundamentals," "patient with kids who have never danced before") that an AI system uses to match a business to a specific query.

A parent asking ChatGPT to recommend a local ballet studio for a six-year-old beginner would not receive a recommendation for this studio. The information needed to make that match did not exist in a form the system could use.

The fix is not technical in a complex sense. It requires consistent identity across all platforms, explicit plain-language descriptions of what the business does and who it serves, active review cultivation with prompts that generate useful language, and content structured around the questions customers actually ask.

The Business Case Is Straightforward

Discovery is the first step in the customer acquisition funnel. If the funnel does not start, nothing downstream matters. Not the quality of the product, not the pricing, not the retention rate once customers enroll.

Small businesses in service categories (studios, clinics, tutoring centers, specialty retailers) are particularly exposed. Their competitive advantage is experiential and relational, qualities that are difficult to surface through transactional digital signals. They are precisely the kind of business that AI recommendation systems struggle to include without sufficient structured corroboration from the web.

The ballet studio grew its enrollment last year. Word of mouth did the work that digital search did not. That is a meaningful outcome. It is also a fragile one. Word of mouth does not scale in the same way that discoverability does. A business that cannot be found by the tools its customers are already using is leaving growth on the table. Not because the product is inadequate. Because the infrastructure for being found has not kept pace with where the searching is actually happening.

The investment required to fix this is modest compared to what most small businesses already spend on web presence management. The cost of not fixing it will compound as AI-mediated discovery continues to replace direct search. A business that cannot be found is not competing. It is waiting to be found by whoever happens to walk by.

Connect with Hitomi on LinkedIn.