A bike shop owner in Zwolle — a mid-sized city in the Netherlands — asked me recently: "Why does my neighbour down the street, with the worse shop, get found by ChatGPT and I don't?" I tested it right away. Typing "best bike shop in Zwolle" into ChatGPT returned three names — his neighbour among them, him nowhere. His Google ranking? Number two, comfortably ahead of that competitor. That is the problem thousands of local business owners run into without realising it: ranking well in Google no longer decides how AI systems make local recommendations.
This article is specifically for you as a local business or small business owner — not for enterprise brands with twenty-person marketing teams. You run a shop, a practice, a restaurant or a service business in a city or region, and you want to show up when someone asks Siri, ChatGPT or Google Gemini for a recommendation "near me".
Why local searches work differently in AI answers
From ten blue links to a single answer
With traditional local SEO you competed for a spot in the Google Local Pack — the three map results with businesses at the top of the search results. Users scanned those options, compared reviews and chose for themselves. In an AI answer, that comparison disappears. Someone asks "which physiotherapist in Groningen specialises in sports injuries" and gets one, maybe two names back — with a short justification. There is no second place to be found in. You are in that answer, or you don't exist for that user.
That sounds harsh, but it is the reality I see every day when testing local queries. AI systems are more selective than search engines, because they have to form an opinion rather than present a list.
What AI models use to find local businesses
ChatGPT, Gemini and Perplexity don't work from their own continuously updated local index the way Google Maps does. They combine multiple sources: real-time search results (via Bing or Google integrations), structured data on your website, and — crucially for local queries — external platforms such as Google Business Profile, Tripadvisor, Yelp-style sites and industry-specific directories. Perplexity and Gemini often use live search results, while ChatGPT with browsing enabled consults similar sources but surfaces local real-time data less consistently.
The upshot: optimising your website alone isn't enough. For local GEO, your ecosystem — your site, your Google Business Profile, your reviews, your listings elsewhere — has to tell one consistent, machine-readable story. For more on how those sources are consulted, see How ChatGPT Finds Your Website (Or Doesn't) — The GEO Guide.
The five building blocks of local GEO
NAP consistency is now an AI trust signal
NAP stands for Name, Address, Phone. It sounds like basic SEO hygiene from 2015, but for AI systems it has become critical. A language model that has to decide whether your business actually exists and is active checks whether your details are identical everywhere: on your website, in your Google Business Profile, on Facebook, in the Chamber of Commerce (KvK) registration that is sometimes factored in. Inconsistencies — an old address on your site, a different phone number on Facebook — are read by AI systems as unreliability, not as carelessness.
In practice: put your NAP details in structured data (Schema.org LocalBusiness markup) on every page of your site, not just the contact page. That is exactly the kind of technical foundation that overlaps with broader bot and crawler accessibility — read more about that in Robots.txt for AI Crawlers: A Practical Guide.
Reviews are the evidence, not the advertising
In traditional marketing, reviews were a trust trigger for human visitors. In GEO they are literally training data and real-time evidence for AI models. When someone asks "which restaurant in Haarlem is good for a business lunch", an AI system doesn't just check that you exist — it scans what people say about you. Specific, recent reviews with concrete details ("quiet atmosphere, good wifi, served within 20 minutes") carry more weight than generic five-star ratings with no text.
Actively ask for reviews with detail after a visit, and respond to every review — including negative ones. Those responses get read too, and they demonstrate active management, which counts as another trust signal.
Local content that actually answers questions
Most local business websites have an "About us" page and a contact page, and that's it. For GEO that is not enough material to draw from. AI systems cite content that answers specific questions: "How much does a driving-licence medical cost in Rotterdam?", "Is parking free at the clinic?", "Is the hairdresser open on Saturdays?".
Write pages or FAQ sections that answer these concrete local questions literally — with prices, opening hours, parking information and specific terms. This is exactly the thinking behind Answer Engine Optimization; the broader approach is set out in Answer Engine Optimization for Businesses: The Practical Guide.
Visibility on the platforms AI actually consults
Your website is one source among many. For local recommendations, AI systems draw heavily on platforms such as Google Business Profile, Tripadvisor, Zorgkaart Nederland (a Dutch healthcare review platform), Werkspot (a Dutch platform for tradespeople) and industry directories. If you invest only in your own website and leave these platforms untouched, you are missing a large share of the sources AI models base their answers on.
At minimum, check every quarter that your listings on these platforms are current, complete and consistent with your website.
A practical step-by-step plan for small businesses without a marketing team
You probably don't have time for a three-month GEO project. Here is what you can pick up this week, ordered by impact versus time invested:
- Test where you stand. Ask ChatGPT, Gemini and Perplexity about your type of business in your town. Are you in the answer? What do they say about you if you are?
- Fix NAP inconsistencies. Compare your website, Google Business Profile and social media. Resolve the differences within a day.
- Complete your Google Business Profile. Opening hours, categories, photos, services — a half-filled profile is a missed opportunity that costs nothing to fix.
- Write three FAQ pages around the questions customers ask you most often by phone or WhatsApp.
- Ask five recent customers this week for a review with concrete details.
These steps need no marketing agency and no technical team — just a few hours of focus. A more extensive version of this kind of quick win, aimed specifically at businesses without a marketing department, is in AEO Without a Marketing Team: 5 Things You Can Do Today.
If you want to take a more structured approach after these first steps, a step-by-step plan tailored to your situation is in GEO Strategy for 2026: A Step-by-Step Guide — the core principles there apply just as well to a sole trader as to a national chain; only the scale differs.
Why this is urgent for local businesses now
Large national chains have marketing budgets to invest on several fronts at once — website, reviews, PR, social. Local small businesses often compete with a handful of direct competitors on the same street or in the same neighbourhood. That means the gap between "visible in AI answers" and "invisible" is often decided by a few simple, cheap actions your competitor hasn't taken yet. Whoever cleans up their NAP data and actively collects reviews now is literally ahead of the bike shop owner who thought his Google ranking was enough.
The full sequence of steps — from verification to the settings that determine your local visibility — is in Creating a Google Business Profile.