If someone types "best HVAC contractor in Philadelphia" into ChatGPT right now, your competitor might get named — and you might not even exist. AI search tools like ChatGPT, Perplexity, and Gemini are already recommending specific local businesses to millions of people, and most small business owners have no idea how it works or whether they're showing up. This guide breaks down exactly what signals these tools use, the four steps to get in front of them, and three specific prompts you can run today to test whether you're showing up right now.
Why AI search visibility is the new SEO (and why it's bigger than you think)
Traditional SEO is about ranking on page one of Google. AI search visibility is something different. When someone asks ChatGPT "who should I hire to fix my AC in San Diego," they don't get ten blue links — they get one or two named businesses and a reason why. That's a winner-take-most situation. If you're the business that gets named, you get the call. If you're not, you don't exist to that person.
I've watched this shift happen fast. A plumber in NJ I worked with told me that three new customers in a single month mentioned they found him by asking ChatGPT. None of them came through Google. He had no idea his name was even appearing in those answers. The visibility was accidental — built on old directory listings and a handful of Google reviews. Now imagine what happens when you build it intentionally.
What GEO is (Generative Engine Optimization)
Generative Engine Optimization — GEO — is the practice of making your business easy for AI systems to find, understand, and recommend. Think of it as SEO, but instead of optimizing for a crawling bot that ranks pages, you're optimizing for a language model that synthesizes information and gives a spoken or written recommendation. The end goal is the same: get your name in front of someone who needs what you do. The mechanics are different.
With traditional SEO, you're chasing keywords and backlinks. With GEO, you're chasing trust signals — structured data, consistent citations, question-answer formatted content, and a clear, authoritative presence across the web. AI tools pull from all of that to decide whose name to say out loud.
How ChatGPT picks businesses to recommend
ChatGPT doesn't have a live connection to every website on the internet. What it does have — especially through its browsing features and the underlying data that trained it — is a strong preference for businesses that are consistently described the same way across many sources, have structured data on their websites, and appear in high-trust directories. When a user asks for a local recommendation, ChatGPT is pattern-matching against everything it knows about businesses in that category and location.
The businesses that get named most often share a few traits: their information is consistent everywhere, they have reviews that include descriptive language about what they do, their websites are written in a way that answers specific questions, and they show up in structured formats that AI can easily parse. None of that is magic — it's just setup work that most small business owners haven't done yet.
The Bing connection most people miss
Here's the piece almost nobody talks about: ChatGPT's live browsing and local data pulls heavily from Bing. Microsoft owns both. When ChatGPT needs real-time local business data, it leans on Bing's index and Bing Places listings more than any other source. That means if you've been ignoring Bing because "nobody uses it," you've been quietly invisible to ChatGPT at the same time.
This one fact changes your whole prioritization. Claiming and completing your Bing Places listing isn't optional anymore — it's one of the most important setup tasks you can do for AI search visibility. I'll walk through exactly how in Step 1 below.
This week: Open Bing Places (bingplaces.com) and search for your business to see whether a listing already exists under your name.
The 6 signals AI tools use to choose your business
AI tools don't make random recommendations. They're pulling on a specific set of signals to decide which businesses to trust. Here's what actually matters.
Bing Places listing
A complete, verified Bing Places listing is the single fastest move you can make for ChatGPT visibility. It tells Microsoft's systems — and by extension, ChatGPT — that your business is real, verified, and located where you say it is. Incomplete listings or missing listings are a dead end for AI recommendations.
NAP consistency across the web
NAP stands for Name, Address, Phone. If your name is "Apex Plumbing LLC" on Google but "Apex Plumbing" on Yelp and "Apex Plumbing Services" on a directory you signed up for in 2017, AI systems see three different entities and trust none of them fully. Consistency is how AI confirms you are who you say you are. Even small differences — "St." versus "Street," suite numbers missing on one listing — create doubt.
Schema markup
Schema markup is structured code you add to your website that tells search engines and AI tools exactly what your business is, where it is, what services you offer, and what questions you answer. Without it, a crawler has to guess. With it, the information is handed over in a format that's easy to read and trust. Most small business websites have zero schema. That's a gap you can close in an afternoon.
Reviews and citations
Reviews matter for two reasons. First, they're content — descriptive text about your business written by real people. When a reviewer writes "best emergency HVAC repair in San Diego, they came same day," that language trains AI to associate you with those specific terms. Second, volume and recency matter. A business with 12 reviews from 2019 reads differently than one with 80 reviews from the last 18 months.
On-page question-answer format
AI tools are built to answer questions. If your website content is written to answer specific questions — "How much does a furnace tune-up cost in Pittsburgh?" or "What's included in a dental cleaning?" — you're giving AI something it can quote directly. Generic service pages that just list what you do give AI nothing to work with. Question-answer formatted content is one of the most underused GEO signals for small businesses.
Topical authority
Topical authority means AI systems recognize you as a credible source on a specific topic in a specific location. You build it by consistently publishing content around your core service area — not about everything, about your thing. A solo attorney in Philadelphia who publishes 15 articles about estate planning in Pennsylvania carries more topical authority on that subject than a general legal blog with 200 posts about every area of law.
This week: Pull up your website's home page and count how many questions it directly answers — most service pages answer zero, and that's the first thing to fix.
Step 1: Claim and complete your Bing Places listing
Why Bing matters for ChatGPT
I said this above but it's worth repeating plainly: ChatGPT uses Bing's local data. When ChatGPT's browsing feature looks up a local service provider, it's pulling from the same index that powers Bing Maps and Bing Search. A missing or incomplete Bing Places listing is the equivalent of being unlisted in the phone book in 1995. You might exist somewhere, but no one can find you.
Google My Business gets all the attention. Bing Places gets almost none. That gap is your opportunity — because most of your competitors haven't touched it either.
The 9 fields that move the needle
When you're inside Bing Places, don't just fill in the basics. These are the nine fields that actually matter for AI visibility:
- Business name — Exact legal or operating name, no extra keywords stuffed in
- Address — Complete, formatted exactly as it appears everywhere else
- Phone number — Local number, same format as on your website and Google
- Website URL — Your actual homepage, not a campaign landing page
- Business category — Primary category as specific as possible (not just "contractor")
- Hours of operation — Including special hours and holiday hours
- Business description — 2-3 sentences that include your city, your core service, and who you serve
- Photos — At least 5, including your storefront or work, your team, and your logo
- Services list — Every individual service you offer, listed by name
This week: Complete all 9 fields in Bing Places — most businesses I check have fewer than 5 filled in.
Step 2: Fix your NAP (name, address, phone) everywhere
Top 12 directories to audit
NAP consistency is one of the most time-consuming parts of GEO setup, but it's non-negotiable. Here are the 12 directories that carry the most weight for local AI visibility:
- Google Business Profile
- Bing Places
- Yelp
- Apple Maps (Apple Business Connect)
- Facebook Business Page
- LinkedIn Company Page
- Better Business Bureau (BBB)
- Angi (formerly Angie's List)
- HomeAdvisor
- Houzz (for contractors and remodelers)
- Justia or Avvo (for attorneys)
- Healthgrades or Zocdoc (for medical/dental)
Pull up each one and compare your business name, address, and phone number character by character. Use a spreadsheet. Write down every discrepancy. This is grunt work, but it pays off.
How inconsistent NAP kills you
AI systems are trying to confirm that you are a real, stable business. When they find your name spelled three different ways across 12 directories, it looks like there might be multiple entities, or like your information is unreliable. Studies have found that citation inconsistency is one of the top factors holding local businesses back from appearing in AI-generated recommendations. One SEO researcher found that correcting NAP inconsistency alone improved local AI visibility for service businesses by as much as 30%.
I've seen this happen with a dental office I helped in NJ — they had five different phone numbers listed across various directories because they'd changed their number twice over the years and never cleaned up the old listings. Once we standardized everything to one consistent NAP, their appearance in local AI results improved within about six weeks.
This week: Run a free Moz Local or BrightLocal scan on your business name to see how many inconsistencies exist before you fix them manually.
Step 3: Add the right schema markup
LocalBusiness schema
LocalBusiness schema tells AI tools and search engines the fundamental facts about your business in a machine-readable format. At minimum, you want your business name, address, phone, URL, business type, geo-coordinates, and hours encoded in your site's code. If your website is on WordPress, plugins like Yoast or RankMath can generate this for you. If you're on a custom build, this goes in the <head> of your site as JSON-LD.
Service schema
Beyond LocalBusiness, add Service schema for each individual service you offer. This tells AI that you specifically provide "emergency furnace repair" or "estate planning for high-net-worth individuals" — in structured, parseable language. The more specific you are in your service schema, the more likely you are to get named when someone asks a specific question. Vague schema produces vague results.
FAQ schema
FAQ schema is one of the most powerful GEO tools available right now and almost no small business websites use it. When you mark up a question-and-answer block on your page with FAQ schema, you're handing AI a direct quote it can use. If someone asks ChatGPT "how much does a furnace replacement cost in Pittsburgh," and your page has FAQ schema with that exact question and a specific answer, you've given the AI a clean, trusted, citable response. That's how you get named.
How to test it
Use Google's Rich Results Test (search.google.com/test/rich-results) and paste in your URL. It will show you every type of schema detected on your page and flag any errors. Also run your site through Schema.org's validator. If the tools come back empty, you have no schema — and that's a gap you can close this week.
This week: Run your homepage URL through Google's Rich Results Test and note what schema, if any, is currently detected.
Step 4: Rewrite your pages to answer questions (not just list services)
The "How do I…" format
Most service pages read like a brochure: "We offer HVAC installation, repair, and maintenance for residential and commercial clients." That's not something AI can quote in response to a user question. Rewrite those same pages to lead with questions: "How do I know if my AC needs to be replaced?" "How much does HVAC installation cost in San Diego?" Answer directly, in plain English, in the first sentence after the question. AI is looking for clean, quotable answers — give it those.
FAQ blocks at the bottom of every service page
Every service page on your site should end with a FAQ block. Five to eight questions that real customers ask you, with direct answers. Not marketing copy. Not "great question, it depends." Specific answers: "A standard furnace tune-up typically costs between $80 and $150 in the Philadelphia area." That kind of specificity is what AI tools are trained to surface. Mark those blocks up with FAQ schema (Step 3) and you've created a machine-readable, AI-friendly content asset on every page of your site.
Direct, quotable answers
The format that works best for AI visibility is what SEOs call a "direct answer snippet" — a 1-3 sentence answer that could stand alone without the rest of the page. Write like you're answering a client on the phone. If someone called you and asked "how long does a roof inspection take," you wouldn't say "it depends on many factors." You'd say "for most single-family homes, a roof inspection takes about 45 minutes to an hour." Write your content that way. AI will find it useful. Your clients will too.
This is also the kind of content work I help clients set up through Apex Autobots — building content systems that auto-draft question-answer page content and FAQ blocks based on the questions their actual customers ask most. It's not magic, it's just a bot that does the repetitive writing work so the business owner doesn't have to.
This week: Pick your top service page and rewrite one paragraph to directly answer the most common question a new customer would have before hiring you.
How to test if you're showing up in ChatGPT
The 3 prompts to try
This is the part nobody else is telling you to do. Before you spend hours optimizing, find out where you actually stand right now. Open ChatGPT (use GPT-4 with browsing enabled if you have access) and run these three prompts, swapping in your own city, service, and business type:
- "Who are the best [your service type] in [your city]?" — Example: "Who are the best estate planning attorneys in Philadelphia?" This is the direct recommendation test. Does your name appear?
- "I need a [your service type] near [your city] — who should I contact?" — Example: "I need a plumber near Pittsburgh — who should I contact?" This simulates the most common conversational AI search query from a potential client.
- "What [your service type] in [your city] has good reviews and is available for [your specific service]?" — Example: "What HVAC contractor in San Diego has good reviews and is available for emergency AC repair?" This tests whether your specific service language and review content is surfacing.
Screenshot your results. If your name doesn't appear in any of these, you have a baseline. Run them again in 60 days after you've completed the four steps above.
What to do if you're not there
If you're invisible in all three prompts, don't panic — most small businesses are. Work through the steps in this article in order: Bing Places first, NAP cleanup second, schema third, content last. Those aren't random — they're ordered by the fastest impact per hour of work. If you want help setting this up, I'm at aaarhontoulis@gmail.com or (484) 602-6390. This is exactly the kind of setup work I do for clients — I can typically have the technical pieces (schema, Bing, NAP audit) sorted within the first week.
How long it takes
Be realistic: this isn't overnight. Bing Places can verify and index within a week or two. NAP corrections across 12 directories take a few weeks to propagate. Schema markup typically gets picked up by crawlers within 2-4 weeks. Content rewrites compound over time. Most clients I work with start seeing their name appear in AI results within 6-10 weeks of completing all four steps. That's not a long time for a channel that could be sending you calls for years. And if you start the 3-prompt test now, you'll have a clear before-and-after to show the impact.
This week: Run all three prompts right now and write down exactly what ChatGPT says — that's your baseline, and you'll want it when your name starts appearing in 8 weeks.
Key takeaways
- AI search tools like ChatGPT are already recommending local businesses to real customers — and most small business owners have no idea whether they're showing up.
- GEO (Generative Engine Optimization) is the practice of making your business easy for AI systems to find, understand, and recommend. It's different from SEO but builds on the same foundation.
- ChatGPT pulls heavily from Bing's data — so a complete Bing Places listing is one of the fastest wins available to you right now.
- The 6 signals that matter most: Bing Places listing, NAP consistency, schema markup, reviews and citations, question-answer formatted content, and topical authority.
- The 4-step setup: claim Bing Places, fix NAP everywhere, add schema markup, rewrite pages to answer questions directly.
- Use the 3-prompt test to find out where you stand today — before you do any of the work — so you can measure the actual impact.
- Most businesses start appearing in AI results within 6-10 weeks of completing this setup. The businesses doing it now will be the ones getting named when their competitors are still wondering why their phone stopped ringing.