How to Rank in ChatGPT Answers

Type “best plumber near me” into ChatGPT or Gemini instead of Google, and you don't get ten blue links to sort through. You get a short, confident answer that names two or three businesses and moves on. If your business isn't one of them, you don't just rank lower, you're invisible for that customer's entire search. That's the shift this guide is about, and it changes what “optimization” even means.

Ranking a page is not the same job as being recommended

Traditional SEO optimizes a page to rank for a query. AI assistants aren't ranking your page at all when they answer a local question, they're answering a question about an entity: a specific business, at a specific place, with a specific reputation. The model has to decide who exists, where, and whether it's any good, then write a sentence about it. Your website is one input into that decision, not the whole decision.

That reframes the work. Instead of asking “how do I get this page to rank,” the useful question is “what would make a model confident enough about my business to name it.” That confidence comes from consistent facts, third-party validation (reviews, citations), and structured signals a machine can parse without guessing.

How AI assistants actually pick who to mention

No AI vendor publishes the exact recipe, and any guide that claims to know the precise weighting is guessing. What the vendors do say, publicly, is more modest and more useful. OpenAI's own help documentation describes ChatGPT choosing to search the web based on what you ask, drawing on live web results rather than a fixed, pre-baked answer (ChatGPT Search, OpenAI Help Center). Google is explicit that its AI Overviews and AI Mode pull from the same underlying Search index as ordinary results, with no separate markup or special optimization required beyond the technical basics every page should already meet (AI Features and Your Website, Google Search Central).

Read plainly, both of those point at the same conclusion: if the underlying web (your site, your directory listings, reviews about you, pages that mention you) is thin, inconsistent, or simply not crawlable, there's less for any AI system to find and trust, regardless of which one it is. Being a well-documented, consistent, genuinely reviewed business is the foundation every one of these tools is built on top of.

The signals that actually move the needle

1. Consistent name, address and phone number (NAP)

If your business is listed as “Aqua Pros Plumbing LLC” on your website, “Aqua Pros” on Yelp, and a P.O. box on an old directory, that's not three sources agreeing, it's three sources a model has to reconcile or discard. Google's own guidance for representing a business is explicit that your listing should reflect your real-world name, location and category (Guidelines for representing your business on Google). Audit your Google Business Profile, your website footer, and the handful of directories you actually appear on, and make them match exactly.

2. A complete, accurate Google Business Profile

Hours, categories, service area, phone number, website link, photos: all of it feeds both classic local search and the AI layer sitting on top of it. An empty or stale profile isn't just a missed opportunity, it's a gap a model has no way to fill in on your behalf. Google publishes the full policy set for what belongs on a profile (All Business Profile policies & guidelines); working through it once, properly, is the single most leveraged hour most local businesses can spend on this.

3. Reviews, and reviews that follow the rules

Review volume, recency and content are a strong proxy for “this business is real and currently operating well,” which is exactly the kind of signal a model leans on when it can't verify a claim itself. Chase genuine reviews from actual customers, respond to them, and don't incentivize or manipulate them: Google's prohibited-content policy specifically calls out paid or traded reviews and unusual review-volume patterns as violations that can get content or an entire profile restricted (Prohibited and restricted content, Google Business Profile Help). A restricted profile helps you nowhere.

4. Structured data that states the facts plainly

Schema.org markup (the LocalBusiness type and its more specific subtypes, like Plumber or Restaurant) puts your name, address, hours and category into a format machines parse directly instead of inferring from prose. Google's developer documentation lays out the required and recommended properties (Local Business structured data, Google Search Central). It won't single-handedly get you named in an answer, but it removes ambiguity that could otherwise cost you the mention.

5. Being cited, described, or linked on pages AI already reads

A model is more willing to name a business other sources already talk about: local press, industry directories, community sites, supplier or partner pages. This is the same logic as old-fashioned citation building for local SEO, and it hasn't stopped mattering just because the search box got smarter. If nobody but you has ever written a sentence about your business, there's little for an AI system to corroborate.

How to test what AI says about you today

You don't need a tool to get a first read. Open a private/incognito window (so history and login state don't skew the answer) in ChatGPT, Gemini, Perplexity and Google, and ask the same two or three questions a real customer would ask: “best [your service] in [your city],” “who should I call for [your problem] near [your neighborhood],” and your own business name to see what it knows about you.

Write down, for each engine: were you named, who was named instead, and what did the answer say about you (if anything). That short log is the baseline you compare against once you've worked through the fixes above.

Common mistakes that quietly cost you the mention

A few patterns show up again and again in businesses that are legitimate and well-reviewed, yet still get skipped. None of them are exotic, and all of them are fixable in an afternoon.

• Duplicate listings. A second Google Business Profile left over from a rebrand, a franchise change, or an old address splits your reviews across two records and leaves a model unsure which one is current.
• A stale review trail. A page of great reviews that stopped a year or two ago reads, to a model weighing recency, like a business that may have closed or changed hands.
• A category mismatch.Listed as “General contractor” when customers actually search “emergency plumber” quietly removes you from the pool being considered in the first place.
• A site a crawler can't read. Content that only renders after client-side JavaScript runs, or a robots.txt rule that blocks it, leaves nothing on your own site for a model to draw on beyond whatever directories say about you.

A starter checklist

• ✓Google Business Profile: fully filled in, correct category, current hours, real photos
• ✓Name, address and phone number match exactly across your site, your GBP listing and every directory you appear on
• ✓A steady, genuine flow of reviews, with no paid or traded reviews
• ✓LocalBusiness (or a more specific subtype) schema markup on your site
• ✓At least a few mentions of your business on sites you don't own: press, directories, partners, community pages
• ✓A private-window test of your top 3 customer questions across ChatGPT, Gemini and Google, repeated monthly

None of this is exotic. It's the same fundamentals that have always underpinned good local visibility, applied with the knowledge that a generative answer now sits in front of the results your customers used to click through. Get the facts right once, keep them consistent, and keep checking what the models actually say.