The question we hear most often from sceptical business owners is some version of this: are people really using ChatGPT to find businesses, or is the whole AI visibility conversation overblown? It is a fair question. There is a lot of noise in the AI space right now, and not every trend is real. So let us look at the evidence, both from the broader market data and from what we are actually seeing in client audits across recruitment, e-commerce, hospitality, professional services, and outdoor businesses on FlinnSchema.
If you would like the broader context on why this matters, our explainer on what AI visibility actually is sets the scene. The short version is that consumer search behaviour is shifting from Google's blue links to AI-generated answers from ChatGPT, Perplexity, Gemini, and Grok. The question is how fast, how broadly, and what to do about it.
The Data: How Big Is the Shift?
The headline figure most commonly cited is Gartner's 2024 forecast that traditional search engine volume will drop 25 percent by 2026 as users move to AI-powered alternatives. That figure has been picked up by countless industry publications, sometimes with caveats and sometimes without. Critics argue the forecast overstates the shift, particularly in older demographics and in regulated industries. Supporters point to OpenAI's growth, with ChatGPT regularly cited as one of the fastest-growing consumer apps in history.
Whichever side you lean toward, the underlying market dynamic is harder to argue with. Consumer search behaviour has changed in measurable ways. Anecdotal but consistent reports from agencies, e-commerce operators, and SaaS founders all point in the same direction: a meaningful share of inbound queries now reference AI tools, and an even larger share of customers say they consult AI before making purchase or service-engagement decisions.
From our own analytics at FlinnSchema, the share of new account signups who mention AI tools in their qualifying questions ("How does this work with ChatGPT?" or "Will my schema help with Perplexity?") has grown sharply over the past twelve months. The category was effectively invisible in early 2025. By mid-2026 it is one of the top three reasons new clients cite for engaging us.
The exact percentage of customers using AI to find businesses is harder to pin down, partly because the same customer often uses both Google and AI, and partly because attribution tracking is poor in AI-driven discovery. What we can say with confidence is that the share is non-trivial, growing, and disproportionately concentrated in younger demographics and in commercial query types.
How Customers Actually Use AI for Discovery
The pattern we see in customer interviews and analytics is that AI is replacing certain search steps rather than replacing search wholesale. A typical AI-driven discovery journey looks like this:
- Initial framing. The customer asks ChatGPT a broad question like "What should I look for in a recruitment agency for tech roles?" or "How do I choose a craft distillery to visit on a Scotland trip?"
- Specific recommendation. Based on the AI's response, the customer narrows their question to "Who are the top recruitment agencies for tech in Kent?" or "Which craft distilleries near Speyside offer tours?" The AI returns specific business names.
- Verification. The customer then often searches those specific business names on Google or directly visits their websites, confirms reviews, checks pricing, and contacts the business.
- Engagement. The actual enquiry or purchase still happens through traditional channels (the business's contact form, phone, or e-commerce checkout).
This is a critical pattern to understand. The AI does not replace the entire funnel. It replaces the discovery step where Google used to dominate. If your business does not appear in step 2, you are out of the consideration set before the customer ever reaches step 3. By the time they are on Google searching for you, they already know your name from the AI recommendation. Businesses that AI engines never mention are systematically excluded from this growing share of discovery, regardless of how well they rank on Google.
Our breakdown of how AI search engines decide which businesses to recommend covers the specific mechanics of step 2 in detail.
What ChatGPT, Perplexity, Gemini, and Grok Actually Return
When you ask one of these AI engines for a business recommendation, the response format and signal weighting differ across engines, but the underlying behaviour is consistent. They issue real-time web searches (this is standard now across all four engines, though Grok and Perplexity were earlier than ChatGPT in mainstreaming it), retrieve relevant pages, extract entity information from those pages, and synthesise a recommendation.
The key thing to understand is that these engines do not have a fixed business directory. They build the response from whatever they find on the web at the time of the query. If your business is not retrievable, structured, and credible at the moment the AI runs its search, you are not in the answer. Our guide on how to get cited by ChatGPT covers the specific changes that move the needle.
The factors that influence whether your business appears in a given answer are well-documented from our 26-factor testing. The highest-impact ones are structured data (especially Organisation, Product or Service, and Review schema), AI crawler access, third-party trust signals, and content that answers customer questions directly in natural language. Pages that read like marketing copy with vague superlatives ("the best", "industry-leading") rarely get cited. Pages that read like helpful explanations with specific claims, prices, addresses, and services are cited far more often.
Real Verticals Where This Is Already Happening
The shift is not happening evenly across industries. Some verticals are already seeing significant AI-driven enquiries, while others are still mostly Google-driven. Based on our client work, the verticals where AI discovery is most established right now are:
- E-commerce, particularly for niche product categories like sustainable goods, craft food and drink, handmade jewellery, and specialty tools. Our e-commerce FAQ guide covers this in detail.
- Local service businesses where customers want a personal recommendation. Our local service FAQ guide covers solicitors, accountants, recruitment agencies, salons, dentists, and others.
- Hospitality and tourism, especially small operators where customers research before booking. We have written specifically about mountain guides and adventure tour operators and small craft distilleries.
- B2B services in specialist niches, where buyers often consult AI before issuing RFPs.
- Professional services like legal, accounting, and consulting, where trust signals matter heavily and AI engines lean on third-party review platforms.
The common thread across these verticals is that customers want a recommendation, not just a list of options. When the goal is "find me the right business", an AI-generated answer is structurally better than a list of ten blue links. That is the channel where AI is most quickly displacing Google.
Case Study: A Recruitment Agency in Kent
One of our long-running clients is a recruitment agency in Kent specialising in tech and engineering placements. When we first audited them in 2025, their AI visibility score was 18 out of 100. They ranked decently on Google for several relevant terms, but in our LLM testing they were mentioned in zero out of forty AI queries. We sent prompts like "Who are the best recruitment agencies for tech roles in Kent?" and "Which UK recruitment agencies specialise in engineering?" to ChatGPT, Perplexity, Gemini, and Grok. The agency did not appear in any answer.
Their issue was not credibility. They had been operating successfully for over a decade with strong client relationships and good Trustpilot reviews. The issue was that AI engines could not parse them. They had no Organisation schema, no Service schema, no LocalBusiness markup, an outdated robots.txt that blocked some AI crawlers, and content written for human marketing rather than AI comprehension.
Over a structured eight-week implementation, we added complete schema markup, opened up crawler access, restructured their key pages to answer common candidate and client questions directly, and consolidated review signals across Google, Trustpilot, and industry-specific platforms. We also added an LLMs.txt file.
By the end of the implementation, their AI visibility score was 62. In our follow-up LLM testing, they appeared in twenty-three out of forty queries across the four engines. ChatGPT and Perplexity cited them by name. Gemini referenced their site as a source. Grok mentioned them in two answers about regional tech recruitment.
The agency reported a noticeable uptick in inbound enquiries within the following quarter. They could not attribute every new lead to AI specifically, but several enquiries explicitly mentioned discovering the firm through ChatGPT. You can find more case study breakdowns on the FlinnSchema results page.
Case Study: A Shopify Jewellery Store
Another representative example is a Shopify store selling handmade jewellery. Their starting AI visibility score was 31. They had eight months of organic growth from Google with reasonable rankings on category terms, but were rarely mentioned in AI-generated answers about gift recommendations or sustainable jewellery brands.
The work we did was almost entirely on the structured data side. Their theme had partial Product schema but missing fields, no AggregateRating, no BreadcrumbList, and an incomplete Organisation block. We added the missing schema types (taking them from two basic types to eight complete types), implemented the FAQ schema on key product pages, opened up AI crawler access in their robots.txt, and added an LLMs.txt file with their brand story and product categories.
Three months later, their AI visibility score was 84. ChatGPT began citing them in answers about sustainable jewellery gifts. Perplexity included them in roundups of UK independent brands. Their Google search impressions also rose by 155 percent in the same period, partly because the schema work has spillover benefits for traditional search rich results.
This is the pattern we see consistently. Structured AI visibility work yields measurable improvements in AI citation rates within weeks to months, and often produces parallel SEO benefits as a side effect. For a step-by-step breakdown of the moves we made, see how to increase your AI visibility score.
Why This Trend Has More Staying Power Than Past Hype
Healthy scepticism about AI hype is warranted. The industry has seen plenty of trends that did not pan out as predicted. So why is AI visibility different?
First, the change is in the search interface itself, not just in the underlying technology. ChatGPT, Perplexity, and Gemini are not faster ways to do Google search. They are a different kind of interface that returns a different kind of answer. Once customers learn the new interface, they tend not to revert to the old one for the same queries. The shift is structural, not novelty-driven.
Second, the supplier side reinforces the trend. Google itself has integrated AI-generated answers into search results (AI Overviews), which means even customers still using Google are receiving AI-generated summaries instead of just blue links. Whether the user thinks of it as ChatGPT or Google, they are increasingly consuming AI-generated content as their primary search output. GEO vs SEO: what changed and what you need to do about it covers the implications of this convergence.
Third, the cost structure favours AI engines. Generating an answer is far cheaper than serving a list of links for query types where the user wants a recommendation rather than research. The economics push more queries toward AI-generated formats over time.
Fourth, attention spans favour AI answers. Customers increasingly want direct answers rather than research tasks. The interface that gives them an answer wins.
What This Means for Your Business
If your customers are in any of the verticals we listed above, or if your customer base skews younger than 45, AI-driven discovery is already shaping a meaningful share of your inbound enquiries. The question is not whether to invest in AI visibility, but how quickly. Businesses that wait twelve more months will be competing against businesses that already have eight to fourteen months of compounding AI visibility work in place.
The practical starting point is to know your number. Our free 26-factor AI visibility audit takes about 60 seconds and gives you a precise score plus a list of the highest-impact gaps. From there, you either implement the fixes yourself or work with our Premium plan for ongoing monitoring, LLM testing, and a prioritised roadmap. For the full anatomy of what we test, see inside the AI visibility audit.
If your AI visibility score is below 30, you are effectively invisible to ChatGPT, Perplexity, Gemini, and Grok for most queries in your industry. That is a channel completely closed to you. Fixing it tends to be cheaper and faster than most SEO programmes, with measurable results within four to eight weeks for the technical work and three to six months for the trust-signal compounding.
If your AI visibility score is between 30 and 60, you are visible but rarely cited. The remaining work is targeted improvement on the specific factors holding you back, which the audit will identify by name.
If your AI visibility score is above 60, you are in good shape and the goal becomes maintenance, expansion to new query types, and defence against competitors who start investing now. For deeper context on what the score itself means, see what the AI visibility score actually means.
The Short Answer
Yes, customers are using ChatGPT and other AI engines to find businesses. The share is non-trivial, growing, and concentrated in commercial query types where customers want recommendations. The verticals most affected today are e-commerce, local services, hospitality, B2B specialists, and professional services. The verticals less affected today are larger-budget enterprise procurement and highly regulated industries with mandated public processes.
Whether your business is in a "yes today" or a "yes eventually" vertical, the structural changes that make AI engines able to find and cite you are the same: complete schema markup, open AI crawler access, third-party trust signals, conversational content, and active maintenance. These investments pay off in AI visibility immediately, in SEO over time, and in customer trust always.
To see exactly where you stand, run a free audit. Or if you would rather discuss your specific business and customer base before doing anything technical, book a free 15-minute walkthrough and we will run live tests on your domain and walk you through what we see.