A homeowner's pipe bursts at 9 PM on a Tuesday. In 2022, they opened Google, typed "plumber near me," scrolled through ten listings, read a few reviews, and called the one with the best combination of stars and availability. The search process took four minutes and produced a shortlist. In 2026, they open ChatGPT and type: "who's the best plumber near me?" ChatGPT responds with one or two names. A direct recommendation, with a reason. The homeowner calls the first one.
The search happened. The result happened. No link was clicked. No list was scrolled. The plumber who got recommended either understood how this works — or got lucky. The plumber who didn't appear either doesn't know this is happening, or hasn't done anything about it yet. This is the moment local business marketing changed.
What's Actually Happening (The Numbers)
This isn't a future trend. It's current behavior at scale. Here's what the numbers look like in 2026:
100M+
ChatGPT queries per day
An estimated 15–20% are local or service-related searches — people asking about businesses, services, and recommendations in specific locations. That's 15–20 million local queries every day going to an AI instead of a search results page.
35–40%
Google searches now show AI Overview
Google AI Overview appears at the top of results for 35–40% of informational and navigational queries. For many local service searches, the AI answer occupies the entire above-the-fold view. A business not in the AI Overview isn't invisible — but it's competing from the second position down.
1–3B
Voice queries answered per day
Siri, Alexa, Google Assistant, and Cortana field between 1 and 3 billion queries daily. Almost every voice result is a single AI-generated answer, not a list. When someone asks their phone "who's the best HVAC company near me?" while driving, they get one name. There is no second place.
The implication is direct: a significant and growing share of searches for local businesses produce an AI-generated answer instead of a list of links. The businesses in those answers win. The businesses not in those answers don't exist for that search. This isn't about being on page 2 of Google — it's about not appearing in a result type that is rapidly becoming the primary search surface.
How AI Decides Which Local Business to Recommend
The recommendation isn't random, and it isn't purely about who spends the most on ads. AI systems pull from structured, consistent, verifiable data. Understanding the signal hierarchy is the first step to doing something about it.
Google Business Profile completeness
GBP is the primary source AI systems use for local business recommendations. Name, address, business categories, services offered, operating hours, service area, photos, and Q&A all feed directly into AI extraction. An incomplete or outdated GBP is the single most common reason a legitimate local business is invisible in AI recommendations.
Review count and recency
More reviews generate more trust signals. Recent reviews signal an active, operating business. AI systems don't just count reviews — they weight recency. A business with 120 reviews from 2022–2023 is weaker in AI signals than a business with 60 reviews where 20 arrived in the last 90 days. Review velocity matters as much as total volume.
Citation consistency across directories
AI systems aggregate reputation and identity data from dozens of sources: Google, Yelp, BBB, Angi, Thumbtack, HomeAdvisor, Apple Maps, and more. When your name, address, and phone number (NAP) match exactly across all of them, AI confidence in your business identity is high. When they conflict — "Suite 100" on Google vs. "Ste. 100" on Yelp vs. a different phone number on HomeAdvisor — AI systems encounter conflicting signals and resolve uncertainty by trusting you less.
Schema markup on your website
Schema markup is machine-readable structured data (JSON-LD format) that explicitly tells AI crawlers what your business is, what it does, where it operates, what it charges, and who its customers are. Without schema, AI systems infer your business data from unstructured text — and inferences are less reliable than declarations. LocalBusiness, Service, and FAQPage schema are the three highest-value types for local businesses.
Answer-ready content structure
AI systems are trained to answer questions. Pages that directly address questions people ask — "How much does AC repair cost in Phoenix?" "What's included in a furnace tune-up?" "How long does water heater replacement take?" — give AI something concrete to extract and cite. Pages that exist only for keyword ranking ("Phoenix HVAC Services | Best HVAC Company Phoenix AZ") provide less extractable value.
Three Local Search Scenarios That Have Changed
Abstract statistics are easy to dismiss. These three scenarios are happening right now, in your market, across every service category.
Scenario 1: Emergency HVAC Call
Homeowner, 10pm, unit stopped cooling
Before (2022)
"HVAC near me" → Google results → homeowner clicks 3 listings, reads reviews on each → calls the one with the best stars and "emergency service available" in the description.
Now (2026)
"Who should I call for emergency HVAC in [city]?" → ChatGPT recommends 1–2 companies by name, with ratings and a note about 24-hour availability → homeowner calls the first one.
What changed:
The middle steps — click, research, compare — are gone. The recommendation is made for the homeowner before they reach your website.
Scenario 2: Attorney Search
Driver injured in accident, needs legal representation
Before (2022)
"Personal injury attorney Charlotte NC" → legal listings on Google plus Avvo → homeowner compares profiles, contingency fee structures, and case types → calls two or three firms.
Now (2026)
"What should I do after a car accident?" → Google AI Overview summarizes the steps — document the scene, see a doctor, contact an attorney — AND recommends attorneys with strong educational content and high citation authority in the same response.
What changed:
The AI answers the informational question AND makes a referral in the same response. Firms with answer-ready content about "what to do after an accident" get the recommendation. Firms without it don't.
Scenario 3: Voice Search in the Car
Homeowner driving home, pipe burst, hands on wheel
Before (2022)
Voice search for "plumber near me" → phone screen shows a list of results → user navigates at a stoplight to pick one and tap the call button.
Now (2026)
"Hey Siri, who's the best plumber near me?" → Siri speaks one answer based on reviews, proximity, GBP data, and schema signals → user says "call them" → done.
What changed:
There is no list. There is one winner. The plumber Siri chose is the only plumber who existed for that search.
What Traditional SEO Gets Right (and What It Misses)
This isn't a case against SEO. It's a case for understanding where the boundaries are.
What traditional SEO gets right:
Quality content — Well-written, structured pages that explain what you do and for whom are more AI-extractable, not less. Content quality is the one investment that pays off for both SEO and AEO.
Backlink authority — High-authority backlinks signal credibility to Google's ranking algorithm. That credibility also influences how AI systems weight your content as a citation source. Links matter on both sides.
Technical site health — Fast load times, clean crawlability, and mobile-friendly design reduce friction for both human users and AI crawlers. A site that loads in under 2 seconds is indexed more reliably.
What traditional SEO misses for AI recommendations:
Keywords in headings and meta titles
These influence Google's ten-blue-links rankings but have minimal effect on whether an AI system recommends your business. AI recommendation engines don't rank you for keyword density.
Schema markup
Most SEO packages don't include LocalBusiness or Service schema implementation. This is precisely the structured data that AI crawlers use to extract and trust your business information. It's not optional for AI visibility.
Citation management
NAP consistency across 40+ directories is an AEO concern, not an SEO concern in the traditional sense. Most SEO engagements never touch citation cleanup.
The practical gap:
A business with excellent traditional SEO — page 1 Google rankings, strong backlinks, well-optimized pages — but no schema markup, no citation management, and an incomplete Google Business Profile will rank well in Google's link results and be invisible in AI recommendations. Both things can be true at the same time. The businesses that close this gap now have a compounding advantage as AI search share grows.
What Local Businesses Should Do in 2026
Four actions. In priority order. The first one is free and takes under an hour.
Claim and complete your Google Business Profile
Every field. Business categories (primary and secondary), services list with descriptions, operating hours including holiday hours, service area boundaries, Q&A section with your most common questions pre-answered, photos of your work and team. GBP is the single highest-leverage action for AI visibility — and it's free. An incomplete GBP is the most fixable problem in local AI search, and the most common one.
Build review velocity
Target 10 or more new reviews per month, consistently. Review count is the primary trust signal across every AI recommendation system. A business with 30 fresh reviews from the last three months signals activity and credibility in a way that 200 older reviews doesn't. The only sustainable way to maintain review velocity is automation: a review request triggered automatically by job completion, sent at the right time, with a direct link. Manual follow-up doesn't scale.
Add schema markup to your website
LocalBusiness, Service, and FAQPage schema are the three highest-value types. This is technical work — JSON-LD code in your site's head section — but the output is that AI crawlers can read your business data precisely instead of inferring it. For most local business websites, a web developer needs 2–4 hours to implement it correctly. The return on that investment compounds every month as AI search share grows.
Audit and fix citation inconsistencies
Check your name, address, and phone number across Google, Yelp, BBB, Angi, Thumbtack, HomeAdvisor, Apple Maps, and Bing Places. Every variation — "Suite 100" vs. "Ste 100," a disconnected old phone number, a previous address still listed on three directories — creates conflicting signals that AI systems resolve by reducing confidence in your data. Consistent NAP across all directories tells every AI system: this business is real, active, and trustworthy.
What This Doesn't Mean
A few things this change is not.
Does this mean SEO is dead?
No. Approximately 42% of searches still result in clicks on organic results, and that traffic still drives real business. A business that ranks well on page 1 of Google for "plumber Charlotte NC" captures homeowners who do still click links — and there are a lot of them. What's changed is that a growing share of searches now bypass the click entirely. Both channels matter. The mistake is optimizing for only one of them.
Do I need to abandon my current marketing to do this?
No. Your Google Business Profile, review management, and content work are already part of good local marketing. The additions required for AI visibility are schema markup — which is technical work your web developer can handle in a few hours — and intentional citation cleanup, which is operational cleanup you can do once and maintain annually. Most of the work extends what you're already doing. You're not rebuilding from scratch.
How fast is this changing?
Faster than mobile did. Mobile search became the primary search surface over about five years — from 2012 to 2017. AI search is making the same transition in two to three years. The businesses that establish AI visibility now get embedded in AI training data and recommendation patterns early. That early positioning compounds: more citations mean more AI appearances, which means more reviews, which means more citations. The window to build that early advantage is the next 12–18 months.
Can I just wait and see how this plays out?
That's a legitimate strategy if your competitors are also waiting. Most aren't. The businesses actively pursuing AI visibility right now are getting recommended in their markets while their competitors don't appear at all. If your primary competitor is already showing up in ChatGPT recommendations for your target service in your city, waiting means you're handing them those leads every day that passes. Check what AI says about your market today — that's the fastest way to know whether waiting is a viable option.
The Bottom Line
Local search is not broken — it's bifurcating. One channel sends homeowners to a list of links and lets them choose. The other sends them directly to a business name and lets them dial. Both channels are active. Both are growing. And right now, most local businesses are optimized for only one of them.
The good news: the four actions that close the AI visibility gap are not complicated. A complete GBP takes an afternoon. Schema implementation takes a developer a few hours. Citation cleanup is methodical, not technically difficult. Review velocity is a system you build once. None of this requires rebuilding your marketing from scratch.
Boxi builds for both traditional SEO and AI search. If you want to understand what your AI visibility looks like right now — and what it would take to appear in ChatGPT and Google AI Overview for your highest-value searches — see the full AEO stack → or book a free audit → We check your AI visibility across ChatGPT and Google AI Overview, audit your schema and GBP, and tell you exactly what to fix.