# Cheers - llms-full.txt ## Version 1.4 (2026-09-02) > Cheers is Y Combinator-backed, done-for-you local visibility for established service companies, especially multi-location brands and PE-backed home-service portfolios. Cheers tracks where each market appears across Google and AI, traces missed recommendations to reviews, profiles, sources, and pages, then helps fix what is holding each market back. This file contains the full text of selected educational articles from the Cheers GEO Academy. For a shorter summary, see /llms.txt. ## Company - **Website:** https://www.cheers.tech - **Contact:** info@cheers.tech | +1 (415) 340-1289 - **Founded:** 2024 | **Backed by:** Y Combinator (S24) - **Founders:** Dylan Allen-Arnegård (CEO), Amadeus Peterson (CTO) ## Solutions - [HVAC](https://www.cheers.tech/solutions/hvac): Seasonal demand, emergency searches, installs, maintenance visits, technicians, comfort advisors, and branch-level reviews. - [Plumbing](https://www.cheers.tech/solutions/plumbing): Emergency repairs, water heaters, drain clearing, leak detection, local trust, service-area reviews, and technician attribution. - [Roofing](https://www.cheers.tech/solutions/roofing): High-ticket projects, final walkthroughs, storm markets, estimator attribution, branch reviews, and AI-cited credibility. - [Franchise Brands](https://www.cheers.tech/solutions/franchise-brands): Brand-wide consistency for salons, med spas, wellness groups, hospitality, and local-service franchises where every location has to earn trust. - [Energy & Smart Home](https://www.cheers.tech/solutions/energy-smart-home): For solar, energy, smart home, and security brands with field sales, installers, technicians, local teams, and high-consideration trust moments. - [Home Service Roll-Ups](https://www.cheers.tech/solutions/home-services): Portfolio-wide reporting for HVAC, plumbing, electrical, pest, garage door, restoration, and franchise service brands. - [Review Generation](https://www.cheers.tech/solutions/review-generation): Always-on review capture with NFC badges, QR flows, employee attribution, tap-to-review reporting, and location scorecards. - [Local Content Generation](https://www.cheers.tech/solutions/local-content-generation): Draft local articles, service pages, and FAQs from real reviews, services, location details, and customer questions. - [Website Optimization](https://www.cheers.tech/solutions/website-optimization): Audit schema, crawler files, page structure, content depth, performance, and service-location coverage so AI and search systems can parse the site. - [AI Visibility Platform](https://www.cheers.tech/solutions/ai-visibility-platform): Prompt tracking, recommendation share, cited-source analysis, competitor mentions, and the local work needed to improve. - [Multi-Location Local SEO](https://www.cheers.tech/solutions/multi-location-local-seo): Location-level visibility across reviews, Google profiles, citations, competitors, service areas, local pages, and operator actions. - [Generative Engine Optimization](https://www.cheers.tech/solutions/generative-engine-optimization-agency): The done-for-you alternative to hiring a generative engine optimization agency: prompts tracked by market, evidence gaps worked every week. - [Multi-Location SEO](https://www.cheers.tech/solutions/multi-location-seo): Multi location SEO run as a managed program: every market measured, worked, and reported on its own, not rolled into one domain number. - [Franchise SEO](https://www.cheers.tech/solutions/franchise-seo): Franchise SEO services that keep every unit findable: profiles, local pages, listings, reviews, and AI answers handled without stepping on the franchisor's brand rules. - [Review Management Services](https://www.cheers.tech/solutions/review-management-services): Done-for-you review management services: capture at the service moment, employee attribution, replies handled, and velocity reported by location. ## Comparisons - [Cheers vs Birdeye](https://www.cheers.tech/compare/cheers-vs-birdeye): An honest Cheers and Birdeye comparison for multi-location service brands: suite breadth versus done-for-you local visibility work, with sources. Reviewed 2026-09-02. - [Cheers vs Podium](https://www.cheers.tech/compare/cheers-vs-podium): Podium converts inbound leads. Cheers wins the local and AI recommendation before the lead exists. An honest comparison with sources and wrong-fit cases. Reviewed 2026-09-02. - [Cheers vs SOCi](https://www.cheers.tech/compare/cheers-vs-soci): SOCi covers localized social, ads, and reputation for enterprise brands. Cheers runs a narrower done-for-you local visibility program. Honest comparison. Reviewed 2026-09-02. - [Cheers vs an SEO agency](https://www.cheers.tech/compare/cheers-vs-seo-agency): Cheers is not an SEO agency. Here is what an agency retainer does better, what Cheers does better, and how to tell which one your locations actually need. Reviewed 2026-09-02. - [Cheers vs LocalClarity](https://www.cheers.tech/compare/cheers-vs-localclarity): LocalClarity publishes per-location pricing for listings and review software you operate. Cheers does the work instead. Honest comparison with sources. Reviewed 2026-09-02. ## Case Studies - [Hello Sugar Case Study](https://www.cheers.tech/proof/hello-sugar): Hello Sugar runs Brazilian wax and sugaring salons in 242 markets. Here is how the brand keeps every one of them visible on Google and in AI answers. - [Sierra Air Conditioning & Plumbing Case Study](https://www.cheers.tech/proof/sierra-cooling): A Las Vegas HVAC and plumbing company with proof anyone can check: open Google, search Sierra Air Conditioning & Plumbing, and count. - [Elite Rooter Case Study](https://www.cheers.tech/proof/elite-rooter): Drain emergencies are won market by market. Elite Rooter tracks reviews, employee cards, and AI answers for all 12 of its markets in one place. - [Action Furnace Case Study](https://www.cheers.tech/proof/action-furnace): An HVAC company with 13,335 Google reviews, 240 employee cards in the field, and attribution data a manager can coach from. ## Key Articles (full text below) - [How does Google AI choose local businesses to recommend?](https://www.cheers.tech/geo-academy/how-local-businesses-can-show-up-in-google-ai-search) - [What Sources Do AI Search Engines Cite?](https://www.cheers.tech/geo-academy/ai-search-engine-source-differences) - [What Is a Good AI Visibility Score?](https://www.cheers.tech/geo-academy/what-is-a-good-ai-visibility-score) - [Best AI Visibility Tools for Local Businesses](https://www.cheers.tech/geo-academy/best-ai-visibility-tools-local-businesses) - [How Much Does AI Visibility Software Cost?](https://www.cheers.tech/geo-academy/how-much-does-ai-visibility-software-cost) - [Best Local SEO Agency Alternatives for Home Services](https://www.cheers.tech/geo-academy/local-seo-agency-alternative-multi-location-home-services) - [Why Does AI Recommend My Competitor?](https://www.cheers.tech/geo-academy/ai-recommends-competitors) - [What Sources Does ChatGPT Use to Give Recommendations?](https://www.cheers.tech/geo-academy/what-sources-does-chatgpt-use) - [What Is Query Fan-Out in Google AI Mode?](https://www.cheers.tech/geo-academy/what-is-query-fan-out-google-ai-mode) - [What Should Location Pages Include for AI Search?](https://www.cheers.tech/geo-academy/what-should-location-pages-include-for-ai-search) - [How Reviews Support AI Visibility for Local Businesses](https://www.cheers.tech/geo-academy/reviews-and-geo) - [What is Generative Engine Optimization (GEO)?](https://www.cheers.tech/geo-academy/what-is-geo) --- ## Hello Sugar Case Study URL: https://www.cheers.tech/proof/hello-sugar Industry: Beauty & Wellness Franchise Published: 2026-06-03 Modified: 2026-06-09 Proof window: February 3, 2025 through June 4, 2026 Summary: Hello Sugar runs reviews, employee cards, and AI visibility checks for 242 salons in one account, and grew its Google review base 85.7% in sixteen months. Proof metrics: - 242 Active local salons: Cheers supports 242 active Hello Sugar salons, plus the parent brand account. Definition: Count of active salon locations under the Hello Sugar brand in Cheers. Source: Cheers product data. Observed: 2026-06-03. Window: Active location footprint. Meaning: Every other number on this page sits on top of this footprint. - 53,578 Active Google reviews: Those salons carry 53,578 Google reviews between them, up from 28,846 in February 2025. Definition: Count of Google Business Profile reviews tied to active Google-connected Hello Sugar locations. Imported-only history is excluded. Source: Cheers product data. Observed: 2026-06-04. Window: Reviews dated April 10, 2015 through June 4, 2026. Meaning: Reviews on the live Google profiles of active connected salons. History imported from other tools is left out of the count. - 274,829 AI visibility checks: Cheers has asked AI engines about Hello Sugar's markets 274,829 times across nine tracked prompts. Definition: Count of AI visibility runs connected to enabled Hello Sugar tracked prompts. Source: Cheers product data. Observed: 2026-06-03. Window: March 9, 2026 through June 3, 2026. Meaning: One check is one live question put to an AI engine about a market Hello Sugar serves. --- ## Sierra Air Conditioning & Plumbing Case Study URL: https://www.cheers.tech/proof/sierra-cooling Industry: HVAC & Plumbing Published: 2026-06-03 Modified: 2026-06-09 Proof window: September 18, 2024 through June 4, 2026 Summary: Sierra grew its public Google profile from 2,503 to 7,088 reviews at a 4.9 rating, and uses AI visibility checks to pick the next fix. Proof metrics: - 7,088 Active Google reviews: Sierra's live Google Business Profile showed 7,088 reviews at 4.9 stars on June 4, 2026. Definition: Live Google Business Profile user rating count for Sierra Air Conditioning & Plumbing. Source: Live Google Business Profile + Cheers product data. Observed: 2026-06-04. Window: Live public profile checked June 4, 2026. Meaning: This is the public number on Google, so you can verify it yourself right now. - +4,585 Google review growth: Sierra had 2,503 connected Google reviews before starting with Cheers in September 2024. Definition: Live Google review count minus the connected Google Business Profile baseline before September 18, 2024. Source: Live Google Business Profile + Cheers product data. Observed: 2026-06-04. Window: September 18, 2024 through June 4, 2026. Meaning: 183.2% growth from the connected baseline in under two years. - 3,720 AI visibility checks: Sierra's tracking covers 10 prompts and 40 prompt-location pairs. Definition: Count of AI visibility runs in the platform-scope tracking used for the Sierra proof story. Source: Cheers product data. Observed: 2026-06-03. Window: May 11, 2026 through June 3, 2026. Meaning: Each run tests how an AI engine answers a service or market question Sierra cares about. --- ## Elite Rooter Case Study URL: https://www.cheers.tech/proof/elite-rooter Industry: Plumbing & Home Services Published: 2026-06-03 Modified: 2026-06-09 Proof window: September 23, 2025 through June 4, 2026 Summary: Elite Rooter tracks reviews, cards, taps, and AI visibility for 12 plumbing markets in one account, and turns the misses into assigned work. Proof metrics: - 12 Active local markets: Elite Rooter operates 12 active local markets in Cheers, plus the parent brand account. Definition: Count of active local markets under the Elite Rooter brand in Cheers. Source: Cheers product data. Observed: 2026-06-03. Window: Active market footprint. Meaning: Each market carries its own reviews, profiles, and AI answers, which is why everything below is tracked per market. - 11,546 Active Google reviews: The active Google-connected markets carry 11,546 reviews, up from 9,836 before September 2025. Definition: Count of Google Business Profile reviews tied to active Google-connected Elite Rooter locations. Imported-only history is excluded. Source: Cheers product data. Observed: 2026-06-04. Window: Reviews dated October 23, 2017 through June 4, 2026. Meaning: Google reviews on active connected markets only. Imported history from other sources is left out. - 16,435 AI visibility checks: Ten tracked prompts across 120 prompt-location pairs have produced 16,435 checks. Definition: Count of AI visibility runs connected to enabled Elite Rooter tracked prompts. Source: Cheers product data. Observed: 2026-06-03. Window: April 29, 2026 through June 3, 2026. Meaning: Enough coverage to compare markets against each other instead of judging the brand on one search. --- ## Action Furnace Case Study URL: https://www.cheers.tech/proof/action-furnace Industry: HVAC Services Published: 2026-06-03 Modified: 2026-06-11 Proof window: November 19, 2025 through June 4, 2026 Summary: Action Furnace runs 240 employee cards against 13,335 Google reviews, with 1,047 reviews traceable to a specific card. Proof metrics: - 13,335 Active Google reviews: Action Furnace's active Google-connected locations carry 13,335 reviews, up from 12,097 before November 2025. Definition: Count of Google Business Profile reviews tied to active Google-connected Action Furnace locations. Imported-only history is excluded. Source: Cheers product data. Observed: 2026-06-04. Window: Reviews dated January 7, 2010 through June 4, 2026. Meaning: Google reviews on active connected locations only. Imported history is left out. - 240 Active employee cards: 240 cards are active in the field, with 2,474 taps recorded. Definition: Count of active employee cards tied to the Action Furnace account. Source: Cheers product data. Observed: 2026-06-03. Window: Cards created December 8, 2025 through May 19, 2026. Meaning: Each card belongs to an employee, which is what makes review attribution possible at all. - 1,047 Card-attributed reviews: 1,047 Action Furnace reviews carry the employee card ID that prompted the review. Definition: Count of Action Furnace review records in Cheers with a non-null employee card ID. Source: Cheers product data. Observed: 2026-06-04. Window: November 19, 2025 through June 4, 2026. Meaning: A review with a card ID can be connected to the employee card the customer used before writing it. - 7,305 AI visibility checks: Seven tracked prompts across 21 prompt-location pairs have produced 7,305 checks. Definition: Count of AI visibility runs connected to enabled Action Furnace tracked prompts. Source: Cheers product data. Observed: 2026-06-03. Window: February 27, 2026 through June 3, 2026. Meaning: Engine-by-engine results, so a miss in ChatGPT and a miss in Google each point to their own fix. --- ## How does Google AI choose local businesses to recommend? Slug: how-local-businesses-can-show-up-in-google-ai-search Category: Playbooks Published: 2026-05-19 Last updated: 2026-09-02 URL: https://www.cheers.tech/geo-academy/how-local-businesses-can-show-up-in-google-ai-search The eligibility requirement Google states for AI features in Search is normal Search eligibility: a page must be indexed and eligible to appear with a snippet. Google's AI Search optimization guide, last updated July 10, 2026, closes off the shortcut most vendors are selling: You don't need to create new machine readable files, AI text files, markup, or Markdown to appear in Google Search. The same guide gives the reason in one sentence: The best practices for SEO continue to be relevant because our generative AI features on Google Search are rooted in our core Search ranking and quality systems. So there is no separate AI file, schema type, or writing trick. Useful location information, accurate Business Profile details, original photos, legitimate reviews, applicable structured data, and a working contact path all help a customer decide, and Google does not require every item on that list. The difference is where the work happens. A multi-location HVAC brand, med spa group, roofing company, pest control franchise, or restoration roll-up cannot optimize only at the corporate website level. AI answers are built from evidence, and local evidence lives at the location level. Plumbing operators can use the industry-specific Google AI Search playbook (https://www.cheers.tech/geo-academy/google-ai-search-playbook-plumbing-companies) to turn this foundation into service-list, location-page, field-proof, and Business Profile decisions. If you want a fast read before you rebuild the source stack, run the Cheers AI Visibility Grader (https://www.cheers.tech/visibility). It gives you a starting snapshot of whether the business appears in AI answers, which sources are showing up, and where a local service brand should inspect first. If your team is debating whether profile cleanup is enough, read Is Google Business Profile Enough for AI Visibility? (https://www.cheers.tech/geo-academy/is-google-business-profile-enough-for-ai-visibility) before treating Google Business Profile as the whole plan. If your team serves customers at their homes, job sites, or properties instead of at a storefront, pair this with How Should Service-Area Businesses Show Coverage for AI Search? (https://www.cheers.tech/geo-academy/how-service-area-businesses-show-coverage-ai-search). Service-area clarity is where profile hygiene, location pages, reviews, and structured data most often drift apart. If the gap is visual proof, use Do photos help AI search recommend local service businesses? (https://www.cheers.tech/geo-academy/do-photos-help-ai-search-recommend-local-businesses) to set a location-level photo standard. When the proof needs motion or spoken explanation, the video SEO guide for home service companies (https://www.cheers.tech/geo-academy/video-seo-home-service-companies) covers watch pages, visible text, media metadata, and location-level measurement. Important: Google does not prescribe a separate AI-optimization checklist. Build useful pages for customers, keep them accessible to Search, and do not claim that every item in this article is a mandatory eligibility requirement. ## What Google AI Search needs from local businesses Google's guide makes five points that matter for local businesses. SEO fundamentals still apply. Google says AI experiences in Search use its existing ranking and quality systems. If your pages are not crawlable, indexable, useful, and technically clean, they are not strong candidates for AI-powered Search features. Useful content beats format tricks. Google specifically pushes back on tactics that exist only for machines, like unnecessary content chunking or AI-only files that do not help users. The page should help the customer first. Non-commodity content matters. Google says sites should create unique value and show why they are a good source. For a local service brand, that means real service expertise, market-specific details, original photos, customer proof, pricing context where appropriate (https://www.cheers.tech/geo-academy/should-home-service-brands-publish-prices-ai-search), license and insurance proof where it changes trust (https://www.cheers.tech/geo-academy/publish-license-insurance-proof-ai-search), and clear explanations of what happens during the job. Business details matter. Google calls out accurate business details as part of AI Search visibility. This is especially important for local brands, where hours, phone numbers, service areas, booking links, categories, and location pages can drift. Use What Should Location Pages Include for AI Search? (https://www.cheers.tech/geo-academy/what-should-location-pages-include-for-ai-search) when the fix needs to happen at the branch-page level. Use Google Business Profile category governance (https://www.cheers.tech/geo-academy/google-business-profile-categories-ai-search) when the profile category itself no longer matches the branch. If after-hours or emergency availability is part of the promise, Where should home service brands publish emergency hours for AI search? (https://www.cheers.tech/geo-academy/where-should-home-service-brands-publish-emergency-hours-ai-search) covers the location-level standard. Agents are coming. Google says site owners should prepare for agentic experiences, where AI assistants help users complete tasks. In local search, that means scheduling, calling, booking, quoting, and comparing businesses. Your website needs to be easy for both people and agents to use. This is unglamorous work. It is harder than publishing one AI article and hoping the model notices. ## Google's AI cites Google, and nobody else does There is a measurable reason to treat Business Profile hygiene as a Google-specific lever. In the Cheers Market Baseline, 32 home-services prompts run across 100 metro service areas on four engines over the 28 days ending September 2, 2026, google.com accounted for 11.4% of the citations on Google's AI Overviews and AI Mode. On the other three engines it was rounding error: 0.2% on ChatGPT, 0.1% on Perplexity, 0.0% on Gemini. Google's AI surfaces were also the ones that named the panel most often, with a 31.6% mean per-organization appearance rate against 25.9% for Gemini and 25.1% for both ChatGPT and Perplexity. Read together, those two numbers say something practical: on Google's AI surfaces, your Google-owned records are load-bearing in a way they are not anywhere else, and a stale profile costs more there than it does in ChatGPT. The engine and metro splits sit in the home services AI visibility index (https://www.cheers.tech/research/home-services-ai-visibility-index-2026), with definitions in the research methodology (https://www.cheers.tech/research/methodology). ## The local translation: AI Search runs on proof For a local service business, "useful content" is not a 2,000-word blog post about what a water heater is. Useful content is the information a customer needs before hiring you. For an HVAC company, that is the boring operational set: which systems you service, whether emergency work is available, response times by market, financing options, maintenance plan details, photos of real installs, technician credentials, recent reviews from that branch, and what happens after someone books. A med spa needs a different set with the same shape. Treatment pages that state contraindications and expectations, provider credentials, the before-and-after policy, location-specific booking links, pricing ranges or consultation details, real patient review themes, and photos of the actual space rather than a stock treatment room. For a franchise service brand, it means every location needs its own evidence. One strong corporate homepage does not prove the Dallas location, the Tampa location, and the Phoenix location all deliver the same experience. AI systems are trying to answer a trust question: "Is this the right business to recommend for this specific customer, in this specific market, for this specific need?" Make that evidence easy to find. ## Query fan-out changes the page strategy Google's guide talks about AI Search using query fan-out, where one user question can trigger many related searches behind the scenes. That matters a lot for local service categories. A customer might ask: "Who should I call for emergency AC repair near Scottsdale?" Behind that one question, the system still has to work out which companies serve Scottsdale, which handle emergencies, which are open right now, which mention AC repair specifically rather than heating and cooling generally, which have recent reviews, which have a phone or booking path that works, and which carry enough reputation proof to be recommended without hedging. That means one generic "HVAC services" page is weak. You need pages and sections that answer the real sub-questions customers ask. Good service pages should cover the work itself, the service area, urgency, process, cost drivers, common failure points, and what makes the business trustworthy. Good location pages should prove that the location is real, active, reviewed, staffed, and available. If the page depends on a scheduler or intake form, pair that work with Can AI Agents Book Appointments From Your Website? (https://www.cheers.tech/geo-academy/can-ai-agents-book-appointments-from-your-website). That is page structure built around proof, not keywords crammed into copy. ## What non-commodity content looks like Google's wording around non-commodity content is important. If your content could appear on any competitor's website with only the logo swapped, it is not doing much work. Local service brands have an advantage here because the business is full of original material: Real jobs. Show the actual work your team performs: installs, repairs, inspections, treatments, service calls, cleanups, restorations, and consultations. Real locations. Mention neighborhoods, service constraints, climate patterns, building types, local regulations, and practical market details. Real people. Show technicians, providers, managers, office teams, trainers, and support staff. AI systems and customers both benefit from clear entity signals. When a named technician or provider affects the buying decision, use the technician bio standard (https://www.cheers.tech/geo-academy/technician-bios-ai-search) to connect that person to a real role, branch, credential, and customer handoff. Real review themes. Do not manufacture review copy. Instead, summarize the themes customers already mention: punctuality, communication, cleanup, professionalism, pricing clarity, follow-through, or bedside manner. If your team needs a repeatable way to turn those themes into page updates, use How to turn customer reviews into AI search content (https://www.cheers.tech/geo-academy/how-to-turn-reviews-into-ai-search-content) before rewriting location pages at scale. Real process. Explain what happens before, during, and after service. Customers want to know what hiring you feels like. If your content sounds like it came from a generic SEO vendor, it is probably not strong enough for AI Search either. ## The technical layer is boring and necessary Google's guide does not make structured data sound glamorous. It should not be glamorous. It should be accurate. For local service businesses, that means: Every location page should be crawlable and indexable, with internal links connecting service pages, location pages, and proof pages. LocalBusiness schema belongs on each location page, Organization schema connects the parent brand to its locations, and Service schema describes the specific work offered. Use sameAs to connect verified profiles such as Google Business Profile, Yelp, Facebook, BBB, Apple Maps, Bing Places, and the industry directories that matter in your category. Review or AggregateRating markup goes in only where it fits current Google guidelines. FAQPage schema is the one to think about twice. Use it only where the FAQ content is visible on the page, and with FAQ rich results retired for most sites in 2026, treat the markup as truthful labeling rather than a result play. Should local businesses still use FAQ schema for AI search? (https://www.cheers.tech/geo-academy/should-local-businesses-use-faq-schema-ai-search) covers the cleanup decision. The goal is not to "trick" AI into understanding you. The goal is to remove ambiguity. This matters even more for multi-location brands. If one location page says "Smith Heating," the Google profile says "Smith Heating and Air - Scottsdale," and Yelp says "Smith HVAC LLC," you are making entity resolution harder than it needs to be. For a deeper cleanup playbook, read How to Fix Conflicting Records Across 50 Locations (https://www.cheers.tech/geo-academy/multi-location-entity-fragmentation). ## The review question: yes, ask customers Review policy gets misunderstood here, so the rule should be clear. You can ask customers for reviews. Google's own review guidance says businesses can remind customers to leave reviews and share a link or QR code. Review requests are normal. They are healthy. Most happy customers will not review unless someone makes it easy. The rules are about how you ask. Do: - Ask every customer through a consistent process - Make the request neutral - Make clear that honest feedback is welcome - Use a direct review link, QR code, or NFC badge - Let customers review later if they prefer - Track which location and employee created the opportunity Do not: - Pay customers for reviews - Offer discounts, gifts, loyalty points, or entries for reviews - Ask only happy customers - Ask for a specific star rating - Ask customers to mention specific words - Pressure customers to review while staff are watching - Set quotas that push employees to solicit a certain number of reviews The rule is plain: asking is good. Pressure is bad. Gating is bad. Customer incentives are bad. Quotas are risky because they can turn a normal ask into pressure. ## How to recognize employees without creating review risk Keep recognition and pay separate from whether a customer posts a review. Google prohibits staff review-count targets, and any reward tied to review volume, rating, or requested content can create pressure or selective asking. NFC badges and QR codes can still provide a consistent, neutral link. The customer chooses whether to use it, what to say, and when to submit. Do not rank employees by link taps, completed reviews, star ratings, or review language. Employee recognition can instead run on verified, job-related evidence: safety and workmanship checks, callback and first-time-fix data, complaint resolution and service recovery, manager quality assurance, policy training and adherence, and specific customer feedback used as context once the facts are verified. Tip: Review attribution can help investigate feedback, but it should not become a review quota or a standalone compensation metric. ## Agent-ready local websites Google's guide talks about agentic experiences. For local businesses, that is not abstract. An agent-ready local website answers a short list without a phone call: what services you offer, where you serve, whether you are open now, whether someone can book online, whether the phone number works, what the service costs or depends on, which location to contact, and what proof exists that this branch is good. If an AI assistant is trying to help a customer book a garage door repair, it should not have to fight your website. Make the conversion path obvious. Use real buttons and links instead of phone numbers baked into images, keep forms simple, label service areas clearly, and put booking and phone actions where a reader expects them. Location pages should carry the exact location name, address, phone number, hours, services, and reviews. None of that is an agent strategy. It is the same page a confused customer needed anyway. ## The 30-day local AI Search plan The practical version is simple. Week 1: Audit the evidence. Pick five priority markets. Search Google, AI Mode if available, ChatGPT, Gemini, and Perplexity for your category. Screenshot who gets recommended and what sources are cited. Compare that to your Google rankings. Week 2: Fix location basics. Check every priority location page, Google Business Profile, Apple Maps listing, Bing Places listing, Yelp profile, BBB profile, and top industry directory. Standardize names, addresses, phones, categories, hours, service areas, and URLs. Week 3: Build better service proof. Add market-specific service content, real photos, process details, provider or technician proof, FAQs, and internal links between service pages and location pages. Week 4: Tighten review operations. Train every frontline employee on the principle of the ask, not a memorized line. Give customers an easy badge, QR, or link path. Track attribution. Reward compliant participation and strong customer experience, not ratings or quotas. Then repeat monthly. AI Search is not a one-time rewrite. It is an operating rhythm. ## What this means Google's AI Search guide is not a rejection of GEO. It is a rejection of shallow GEO. For local service businesses, the strategy is not mysterious. Make the website useful and crawlable, make every location easy to verify, keep business details accurate, publish proof customers can trust, use structured data carefully rather than liberally, ask every customer for a review through a neutral process, attribute review activity to the frontline, reward the behavior instead of the rating, and watch AI recommendations market by market. That is the operating rhythm. If you want to see how AI currently reads your business, run the free AI Visibility Grader (https://www.cheers.tech/visibility). ## Sources Checked September 2, 2026. - Optimizing your website for generative AI features on Google Search. Google Search Central, last updated July 10, 2026. developers.google.com/search/docs/fundamentals/ai-optimization-guide (https://developers.google.com/search/docs/fundamentals/ai-optimization-guide). Source of both quoted sentences, the five points, and the agentic-experience guidance. - Google Search Essentials. Google Search Central, accessed September 2026. developers.google.com/search/docs/essentials (https://developers.google.com/search/docs/essentials). Source of the eligibility baseline. - Search Essentials: technical requirements. Google Search Central, accessed September 2026. developers.google.com/search/docs/essentials/technical (https://developers.google.com/search/docs/essentials/technical). Source of the crawlability and indexable-content requirements. - Crawling and indexing documentation. Google Search Central, accessed September 2026. developers.google.com/search/docs/crawling-indexing (https://developers.google.com/search/docs/crawling-indexing). Source of the crawlable-link and robots-control guidance. - Local business (LocalBusiness) structured data. Google Search Central, last updated December 10, 2025. developers.google.com/search/docs/appearance/structured-data/local-business (https://developers.google.com/search/docs/appearance/structured-data/local-business). Source of the LocalBusiness markup guidance. - Review snippet (Review, AggregateRating) structured data. Google Search Central, accessed September 2026. developers.google.com/search/docs/appearance/structured-data/review-snippet (https://developers.google.com/search/docs/appearance/structured-data/review-snippet). Source of the current Review and AggregateRating rules. - Tips to get more reviews. Google Business Profile Help, accessed September 2026. support.google.com/business/answer/3474122 (https://support.google.com/business/answer/3474122?hl=en). Source of the permission to remind customers and share a review link. - Prohibited and restricted content. Google Maps User Generated Content Policy Help, accessed September 2026. support.google.com/contributionpolicy/answer/7400114 (https://support.google.com/contributionpolicy/answer/7400114?hl=en). Source of the incentive, selective-solicitation, and staff-target prohibitions. - Read and reply to reviews. Google Business Profile Help, accessed September 2026. support.google.com/business/answer/3474050 (https://support.google.com/business/answer/3474050?hl=en). Source of the review-response guidance. - A new era for AI in Search. The Keyword, Google, 2026. blog.google/products-and-platforms/products/search/search-io-2026 (https://blog.google/products-and-platforms/products/search/search-io-2026/). Source of the AI Mode, AI Overviews, and agentic local feature announcements. - Local Consumer Review Survey 2026. BrightLocal, 2026. brightlocal.com/research/local-consumer-review-survey (https://www.brightlocal.com/research/local-consumer-review-survey/). Source of the local AI discovery and review verification behavior. - Cheers Market Baseline and panel: 32 home-services prompts across 100 metro service areas on four engines, plus 119 home-services organizations, 28 days ending September 2, 2026, aggregate only. Home services AI visibility index (https://www.cheers.tech/research/home-services-ai-visibility-index-2026) and methodology (https://www.cheers.tech/research/methodology). Source of the google.com citation shares and per-engine appearance rates. *Dylan Allen-Arnegård (https://www.linkedin.com/in/dylan-allen/) is the CEO of Cheers, the local search platform for service businesses.* --- ## What Sources Do AI Search Engines Cite? Slug: ai-search-engine-source-differences Category: Research Published: 2026-03-20 Last updated: 2026-09-02 URL: https://www.cheers.tech/geo-academy/ai-search-engine-source-differences Four engines, four different source diets. In the 28 days ending September 2, 2026, the Cheers market baseline ran 32 home-services prompts across 100 metro areas and collected 356,019 citations. Gemini pointed at the contractor's own website in 80.7% of its citations. ChatGPT did that in 48.8%, and sent 44.7% to directories and review platforms instead. Same question, same metros, same week. That is the gap this article is about: the documentation tells you what each engine can do, and the citation counts tell you what it actually did. Last verified: September 2, 2026. That is the comparison local operators can defend from the engines' own documentation as of September 2, 2026. ## ChatGPT Search OpenAI's help article on searching the web, updated in late August 2026, says ChatGPT search rewrites your question into targeted queries and sometimes hands those to other search providers, sharing an approximate location derived from your IP address so that "restaurants near me" becomes something like "top restaurants San Francisco". It is explicit that nothing about the result is owed to you: ChatGPT ranks search results using multiple factors intended to help users find relevant, reliable information. Placement is not guaranteed. The same page names actual partners in one narrow case: for restaurant reservations, availability may come from OpenTable, Resy, or Yelp. Outside that, OpenAI does not promise that a Google Business Profile, a Yelp page, or any other directory feeds a local answer. It also tells site owners the one concrete eligibility step, which is allowing OAI-SearchBot to crawl the site. ## Gemini and Google AI Mode Gemini Apps and AI Mode are separate surfaces. Google's Gemini help page is direct about how thin the source trail can be: Not all responses include related links or sources. When links do appear, Google says they may include public websites, files you uploaded, or Workspace documents and email if you connected Workspace. A missing Sources button is not evidence that nothing was retrieved. For AI Mode, Google's Search guide says the feature retrieves from the core Search index and can issue related searches through query fan-out. The response shows supporting web links. Google's local guidance tells site owners to keep Business Profile information current. That supports treating the profile and the branch website as source truth. It does not establish a fixed citation weight for either one. ## Perplexity Perplexity searches the live web and attaches source links to the answer. Its source-label article, last updated August 7, 2026, describes three domain labels, Government, Academic, and Trusted, and then limits what they mean: A label describes the website as a whole, not any single article or claim on it. Perplexity also says most domains carry no label at all, and that the absence of one is not a negative judgment. For a plumbing company, no label is the normal state. The useful local audit is the answer's actual source list, because Perplexity publishes no rule that one review site or directory controls local recommendations. ## Microsoft Copilot Microsoft says Copilot can turn a prompt into a shorter Bing query, ground the response in Bing results, and show the exact query and sources. Bing says its index is built from crawled third-party pages and that business queries may return details such as store hours and location. For local visibility, make sure Bing can reach the correct branch page and that the public business facts do not conflict. Do not assume that a strong Google result automatically creates the same Copilot result. ## What the citation mix actually looks like Documentation says what an engine can do. Counting citations says what it did. Yext analyzed more than 6.8 million citations across 1.6 million responses in October 2025 and found the engines disagree about what a good source is. 52.15% of Gemini citations came from brand-owned websites. 48.73% of ChatGPT citations came from third-party sites such as Yelp, TripAdvisor, and MapQuest. On subjective queries, directory citations inside ChatGPT rose to 46.3%, and niche industry sources made up 24% of Perplexity's citations. The Cheers market baseline runs a narrower version of that test on home services only: 32 natural-language trade prompts, 100 metro service areas in the United States and Canada, four engines, 28 days ending September 2, 2026, 356,019 citations. Every cited domain is sorted into contractor sites, directories and review platforms, social and forums, media and vendors, or Google itself. Aggregate only, no customer named. The methodology (https://www.cheers.tech/research/methodology) and the wider cuts are in AI visibility statistics (https://www.cheers.tech/research/ai-visibility-statistics). Gemini sent 80.7% of its citations to a contractor website and Perplexity 72.3%, so on those two surfaces a thin or unreachable location page costs the citation outright. ChatGPT split almost evenly, 48.8% contractor sites against 44.7% directories and review platforms, which means a stale Angi or BBB record is doing real work there. Google AI Overviews and AI Mode is the only engine where google.com itself is a meaningful share, at 11.4%. The most cited platform domains across all 100 metros were google.com, expertise.com, angi.com, bbb.org, and reddit.com. That is a list worth checking your own records against before deciding a directory does not matter. ## What I audit across all four I save the prompt, engine, location context, businesses named, pages cited, and time checked. Then I group the sources into the business website, platform-managed business records, review or directory pages, publishers, and other third-party evidence. The provider-tracking guide (https://www.cheers.tech/geo-academy/best-platforms-track-chatgpt-gemini-perplexity) covers the tools that preserve that record. The citation cleanup guide (https://www.cheers.tech/geo-academy/citations-stack-2025) covers conflicting branch facts once the source audit finds them. I do not tell a client to build a listing because a generic checklist says an engine likes it. I want to see that source appear for the actual service, market, and competitor set, or know that customers already depend on it. The Cheers AI Visibility Grader (https://www.cheers.tech/visibility) runs the first version of that check on one profile across ChatGPT, Gemini, and Perplexity. Crawler permissions and optional source maps are separate from citation measurement. What Is LLMs.txt? (https://www.cheers.tech/geo-academy/what-is-llms-txt) explains what that file can and cannot do. ## Sources Every link below was opened and checked on September 2, 2026. - Searching the web with ChatGPT (https://help.openai.com/en/articles/9237897). OpenAI Help Center, updated August 2026. Search behavior, query rewriting, search partners, approximate location, reservation providers, ranking, and OAI-SearchBot eligibility. - View related sources in Gemini Apps (https://support.google.com/gemini/answer/14143489?hl=en). Google Gemini Apps Help. Which sources related links can include and when a response has none. - Optimizing your website for generative AI features on Google Search (https://developers.google.com/search/docs/fundamentals/ai-optimization-guide). Google Search Central, last updated July 2026. AI Mode retrieval, query fan-out, and business-detail guidance. - Understanding source labels (https://www.perplexity.ai/help-center/en/articles/20260806-understanding-source-labels). Perplexity Help Center, updated August 7, 2026. What the Government, Academic, and Trusted labels do and do not mean. - Practical tips for using Perplexity (https://www.perplexity.ai/help-center/en/articles/10352971-practical-tips-for-using-perplexity). Perplexity Help Center. Live-web search and source links. - How web search works in Microsoft 365 Copilot (https://support.microsoft.com/en-us/microsoft-365-copilot/how-web-search-works-in-microsoft-365-copilot-chat-and-agents). Microsoft Support. Bing query generation, grounding, and the Sources control. - How Bing delivers search results (https://support.microsoft.com/en-us/bing/how-bing-delivers-search-results). Microsoft Support. Bingbot crawling, index construction, and business results such as store hours and location. - AI Visibility in 2025: How Gemini, ChatGPT, and Perplexity Cite Brands (https://www.yext.com/blog/ai-visibility-in-2025-how-gemini-chatgpt-perplexity-cite-brands). Yext, October 2025. 6.8 million citations across 1.6 million responses. - Cheers market baseline, 28 days ending September 2, 2026 (https://www.cheers.tech/research/ai-visibility-statistics). 32 home-services prompts, 100 metro areas, four engines, 356,019 citations, aggregate only. Methodology (https://www.cheers.tech/research/methodology). *Dylan Allen-Arnegard (https://www.linkedin.com/in/dylan-allen/) is the CEO and Co-Founder of Cheers. I help multi-location service brands get recommended by AI search and Google, with results tracked by location, employee, and competitor.* --- ## What Is a Good AI Visibility Score? Slug: what-is-a-good-ai-visibility-score Category: Research Published: 2026-05-29 Last updated: 2026-09-02 URL: https://www.cheers.tech/geo-academy/what-is-a-good-ai-visibility-score A good local AI appearance rate is 37.3 percent or higher. That is the top-quartile line in the Cheers panel, 28 days ending September 2, 2026, 119 home-services organizations, aggregate only. The median is 27.4 percent, and 47.4 percent marks the top decile. Last verified: September 2, 2026. My benchmark is: - Below 15.1 percent: bottom quartile. The business is absent from most tracked buying prompts. - Around 27.4 percent: median. The business appears in fewer than a third of tracked answers. - 37.3 percent or higher: good. This is top-quartile performance in the panel. - 47.4 percent or higher: top decile. The business appears in nearly half of tracked answers. The panel mean is 27.0 percent and the mean citation rate is 22.0 percent. Every organization in the panel tracks roughly 31 buying prompts, the window is the 28 days ending September 2, 2026, and the only figures I publish are aggregates: no single organization's results are broken out, and the panel is home services, so a dentist or a law firm should not read these bands as their own. The full cuts are in AI visibility statistics (https://www.cheers.tech/research/ai-visibility-statistics), the sampling rules in the methodology (https://www.cheers.tech/research/methodology), and the AI visibility benchmark report (https://www.cheers.tech/geo-academy/ai-visibility-gap) explains the underlying recommendation gap. ## One brand, four different scores The same 119 organizations score differently depending on which engine you ask. Median appearance by engine, same window: Google AI Overviews and AI Mode 31.8 percent, Gemini 25.6 percent, Perplexity 23.9 percent, ChatGPT 21.6 percent. Citation rates diverge harder than that. The panel mean citation rate is 34.3 percent on Perplexity and 12.4 percent on Gemini, so a brand can be named often on one surface and almost never used as a source on another. Prompt shape moves the number too. Price and cost questions returned a 37.9 percent pooled appearance rate across the panel; "best near me" and named-place prompts returned 22.8 percent. If a vendor quotes you one score, ask which engines and which prompt shapes are behind it before you compare it with anything. When ChatGPT did name a panel organization, the mean mention position was 2.87. Third on the list is a different commercial outcome from first, and an appearance rate on its own will not show you that. ## Use appearance rate before a blended score Vendors calculate AI visibility in different ways, and they say so in their own documentation. BrightLocal describes its visibility score like this: It combines how often you're mentioned, where you appear in answers, how positively you're described, and how strongly you're recommended to a score out of 100. Local Falcon scopes its metric far more tightly: Share of AI Voice (SAIV) is a metric that shows how visible your brand is within AI-generated search results for a specific query and area. One is a blended index out of 100 across platforms. The other is a per-query, per-area share. Both are useful inside their own tool. Neither converts into the other, and a vendor quoting "your AI visibility score went up 12 points" has told you nothing you can audit. I start with appearance rate: the share of tracked answer runs in which the business appears. It is plain enough to audit from the saved responses. Citation rate answers a separate question: how often the business or its site is cited as a source. ## A score without location detail misleads Suppose a parent brand has a healthy aggregate because a few mature branches appear often. That number can hide a new acquisition whose name, page, Business Profile, reviews, and citations still disagree. The correct operating view breaks the score down by location, buyer prompt, engine, competitor, and cited source. A plumbing branch can perform well for drain cleaning and disappear for water-heater replacement. An HVAC branch can appear in Gemini and miss in ChatGPT. The average does not tell the owner which problem to fix. The location-level audit guide (https://www.cheers.tech/geo-academy/how-to-audit-ai-search-visibility-across-locations) shows how to store those dimensions. The source comparison (https://www.cheers.tech/geo-academy/ai-search-engine-source-differences) explains why one engine cannot stand in for the rest. ## How to use the percentile bands Treat the panel bands as an operating benchmark, not a promise. If a branch is below 15.1 percent, inspect whether the prompts match real services and whether the public facts are correct before starting content work. If it is near the median, compare lost prompts with the competitors and sources that appear instead. If it is above 37.3 percent, protect the working source coverage and look for weak service lines rather than chasing the aggregate. Re-run the same prompt set on a consistent cadence. Changing the engines, markets, or questions changes the denominator and breaks the trend. You can establish a current baseline with the AI Visibility Grader (https://www.cheers.tech/visibility). For a multi-location program, keep the branch-level detail even when leadership receives one rollup. ## Sources Every link below was opened and checked on September 2, 2026. - Cheers panel, 28 days ending September 2, 2026 (https://www.cheers.tech/research/ai-visibility-statistics). 119 home-services organizations, aggregate only, about 31 tracked buying prompts per organization. Supports the appearance-rate mean, the percentile bands, the per-engine medians, the prompt-shape rates, and the citation-rate means. Figures pulled September 3, 2026. Methodology (https://www.cheers.tech/research/methodology). - What does each Local AI Visibility metric mean? (https://help.brightlocal.com/hc/en-us/articles/37990220496018-What-does-each-Local-AI-Visibility-metric-mean). BrightLocal Help Center. Definitions of visibility score, mention rate, share of voice, and average position. - Local Consumer Review Survey 2026 (https://www.brightlocal.com/research/local-consumer-review-survey/). BrightLocal, February 2026. Consumer-side adoption context for why the appearance rate matters. - How to use Share of AI Voice (https://www.localfalcon.com/knowledge-base/kb92-how-to-use-share-of-ai-voice-saiv). Local Falcon Knowledge Base. The per-query, per-area definition. - Optimizing your website for generative AI features on Google Search (https://developers.google.com/search/docs/fundamentals/ai-optimization-guide). Google Search Central, last updated July 2026. What Google says eligibility for AI surfaces depends on. *Dylan Allen-Arnegard (https://www.linkedin.com/in/dylan-allen/) is the CEO and Co-Founder of Cheers. I help multi-location service brands get recommended by AI search and Google, with results tracked by location, employee, and competitor.* --- ## Best AI Visibility Tools for Local Businesses Slug: best-ai-visibility-tools-local-businesses Category: Comparisons Published: 2026-06-13 Last updated: 2026-09-02 URL: https://www.cheers.tech/geo-academy/best-ai-visibility-tools-local-businesses The best AI visibility tools for a local or multi-location operator are Cheers, Local Falcon, BrightLocal, Otterly, Semrush, Ahrefs Brand Radar, Peec AI, and Profound. The right choice depends on whether you need a software report or a team that fixes the local inputs behind it. Last verified: September 2, 2026. My criteria are simple: the tool must track real buyer prompts, preserve the cited sources, separate locations, show the local competitors, and make the next owner obvious. A local score without those details is hard to use. Prompt shape decides what you see. In the Cheers production panel covering the 28 days ending September 2, 2026, pooled across 119 home-services organizations, price and cost questions produced a 37.9% appearance rate while "best near me or in a named place" prompts produced 22.8%. A tool loaded only with map-pack style prompts will report a lower number for the same business. Aggregate only, no organization named; see the panel methodology (https://www.cheers.tech/research/methodology) and AI visibility statistics (https://www.cheers.tech/research/ai-visibility-statistics). ## The local operator shortlist - Cheers: done-for-you work for multi-location service brands, with tracking by location, prompt, engine, competitor, source, and employee-linked review operations. - Local Falcon: self-serve geo-grid scanning for supported AI and map surfaces, built around a business location. - BrightLocal: local SEO software with Local AI Visibility included on active locations in its Track, Manage, and Grow plans. - Otterly: lower-cost self-serve prompt and citation monitoring. Its plan page lists "Tracking of 4 AI Search Engines: ChatGPT, Google AI Overviews, Perplexity, MS Copilot" in the base, with "Claude, Google AI-Mode, Gemini as extra Add-ons." - Semrush: AI visibility inside a broader SEO workflow, sold per domain with custom prompts. - Ahrefs Brand Radar: broad brand, web, and AI monitoring with custom prompts that retain location context. - Peec AI: model-level visibility, citation, sentiment, and competitor analytics for marketing teams. - Profound: brand and content monitoring from a ChatGPT-only entry plan at $99 per month billed yearly to Enterprise coverage of up to nine answer engines. ## Start with the location, not the brand average A home-services company does not win one national answer. It wins or loses an emergency plumbing question in one market, a heat-pump question in another, and a roofing comparison somewhere else. That makes the reporting unit important. Local Falcon uses geo-grid scans, and BrightLocal organizes Local AI Visibility inside an active location. Semrush can generate prompts from a domain and location, Ahrefs carries location context into custom prompts, and Peec supports projects and countries. Profound publishes region limits by plan; Otterly supports multiple countries and workspaces. Cheers goes further for the narrow market we serve. We connect each missed recommendation to the location page, Business Profile, review pattern, citation, or local source that needs work, then we own that work with the client. That is a service model, not a lighter software plan. ## Where broad platforms fit Broad AI visibility platforms are useful when the business wants to monitor brand mentions, corporate narratives, PR sources, content gaps, and market perception. The right question is fit, not which logo has the longest feature list. Use this buyer shortlist by job. Cheers appears in that comparison because I want the boundary to be clear. We are not the right choice for corporate PR monitoring, an ecommerce catalog, or a single-location team that only wants to inspect raw answers. Ahrefs, Semrush, Profound, Peec, or Otterly will usually be cleaner for those jobs. If the core need is a local geo-grid, start with Local Falcon. If your team already lives in BrightLocal, its active-location view can reduce tool sprawl. If nobody on your side has time to turn findings into page, profile, review, listing, and content work, a self-serve monitor will document the problem without fixing it. ## How I would evaluate a trial Run the same service question in at least two markets. Confirm the exact engine, location setting, answer, competitors, cited pages, refresh cadence, export options, and plan limit. Then ask who owns the first corrective action. Do not accept a sales demo built only around the parent brand name. A buyer is more likely to ask who can solve a specific problem in a specific place. Your test set should match that decision. Whatever the tool reports, the fix usually lands on surfaces Google documents. Its AI features guidance, last updated December 2025, is short about the eligibility question: "There are no additional technical requirements." The work is the underlying page, profile, and review evidence. Use the AI Visibility Grader (https://www.cheers.tech/visibility) for a baseline. Then follow the location-level audit workflow (https://www.cheers.tech/geo-academy/how-to-audit-ai-search-visibility-across-locations) if you need a repeatable record across branches, and the price anchors (https://www.cheers.tech/geo-academy/how-much-does-ai-visibility-software-cost) when the shortlist is down to two. ## Sources - Local Falcon AI scan reports (https://www.localfalcon.com/knowledge-base/kb96-how-to-read-an-ai-visibility-scan-report). Official location, geo-grid, competitor, and source details checked September 2, 2026. - BrightLocal Local AI Visibility (https://help.brightlocal.com/hc/en-us/articles/37989673412114-What-is-Local-AI-Visibility). Official active-location, platform, prompt, competitor, and source details checked September 2, 2026. - Otterly pricing (https://otterly.ai/pricing). Official engine, add-on, workspace, and prompt details checked September 2, 2026. - Semrush AI visibility features (https://www.semrush.com/kb/1626-ai-visibility-features). Official domain, location, engine, prompt, and update details checked September 2, 2026. - Ahrefs Brand Radar (https://help.ahrefs.com/en/articles/11064852-what-is-brand-radar-and-how-to-use-it). Official custom-prompt, location, engine, and price details checked September 2, 2026. - Peec AI pricing (https://peec.ai/pricing). Official model, project, country, and plan details checked September 2, 2026. - Profound pricing (https://www.tryprofound.com/pricing). Official engine, prompt, region, and plan details checked September 2, 2026. Starter tracks ChatGPT only; Growth tracks three answer engines; Enterprise covers up to nine. - AI features and your website (https://developers.google.com/search/docs/appearance/ai-features). Google Search Central, last updated December 2025. Checked September 2, 2026. - AI visibility statistics (https://www.cheers.tech/research/ai-visibility-statistics) and methodology (https://www.cheers.tech/research/methodology). Cheers Research, September 2, 2026. Appearance rates by prompt shape across 119 home-services organizations, 28 days ending September 2, 2026. *Dylan Allen-Arnegard (https://www.linkedin.com/in/dylan-allen/) is the CEO and Co-Founder of Cheers. I help multi-location service brands get recommended by AI search and Google, with results tracked by location, employee, and competitor.* --- ## How Much Does AI Visibility Software Cost? Slug: how-much-does-ai-visibility-software-cost Category: Comparisons Published: 2026-06-14 Last updated: 2026-09-02 URL: https://www.cheers.tech/geo-academy/how-much-does-ai-visibility-software-cost AI visibility software starts at $24.99 per month for Local Falcon, $29 per month for Otterly, $29 per month billed annually for BrightLocal Track, and $99 per month for Semrush or Profound. Ahrefs Brand Radar starts at $199 per month for one AI index. Peec publishes plan limits, but its reviewed pricing page did not expose a dollar amount. Those prices were checked on each vendor's official page on September 2, 2026. They are not equivalent plans. Last verified: September 2, 2026. ## Public price anchors Local Falcon: Starter is $24.99 per month for 7,500 credits. AI visibility tracking is listed in the plan matrix. Credits are spent on geo-grid data points. Otterly: Lite is $29 per month for 15 prompts. Standard is $189 per month for 100 prompts. Premium is $489 per month for 400 prompts. The plan page lists "Tracking of 4 AI Search Engines: ChatGPT, Google AI Overviews, Perplexity, MS Copilot" in the base plans, with "Claude, Google AI-Mode, Gemini as extra Add-ons." BrightLocal: Track is $29 per month when billed annually for one location. Local AI Visibility is included for active locations on Track, Manage, and Grow plans. Semrush: AI Visibility costs $99 per month per domain when billed annually. The base plan includes 25 custom prompts and mentions from ChatGPT, Google AI, Gemini, and Perplexity. Profound: Starter is $99 per month billed yearly for ChatGPT and 50 prompts. Growth is $399 per month billed yearly for three answer engines and 100 prompts. Enterprise is custom. Ahrefs Brand Radar: the help page prices it as "Single platform: $199/month per index" and "All platforms: $699/month (includes 2,500 custom prompt checks/month)." Ahrefs also publishes separate custom-prompt tiers between $50 and $250 a month. Peec AI: the public page lists Starter, Pro, Advanced, and Enterprise limits, plus a model list and a 15 percent annual discount. It did not expose a dollar amount in the reviewed page, so I am not repeating a price from a secondary source. Add-on pricing is where a cheap plan gets expensive, and the reason is measurable. In the Cheers production panel covering the 28 days ending September 2, 2026, across 119 home-services organizations, the mean per-organization citation rate was 34.3% on Perplexity and 28.7% on Google AI surfaces but only 13.4% on ChatGPT and 12.4% on Gemini. Buying a single-engine plan means measuring one of four very different pictures. Aggregate only, no organization named; see the panel methodology (https://www.cheers.tech/research/methodology) and AI visibility statistics (https://www.cheers.tech/research/ai-visibility-statistics). The table keeps the published unit beside the price. A prompt, a domain, a model, a geo-grid credit, and an active location are different units. Price comparisons that ignore the unit are useless. ## How Cheers is priced Cheers has no public list price. We price per location based on the operating scope. Cheers is a done-for-you service that manages the website, reviews, listings, structured data, local content, and AI visibility program for multi-location service businesses. That means the fair comparison extends beyond the monitoring subscription. Include the internal or agency labor required to inspect citations, assign work, update the source, coordinate branches, and verify the result after the change. Cheers is not the right choice when a team only wants raw monitoring software or a one-time brand check. The self-serve tools above cost less because the buyer owns the response to the report. Use the demo action below if you need a per-location scope for a managed program. For software-only fit, use the platform comparison (https://www.cheers.tech/geo-academy/best-platforms-track-chatgpt-gemini-perplexity) and the local buyer shortlist (https://www.cheers.tech/geo-academy/best-ai-visibility-tools-local-businesses). The managed alternative is the AI visibility platform (https://www.cheers.tech/solutions/ai-visibility-platform) service line. ## Sources - Local Falcon pricing (https://www.localfalcon.com/pricing). Official credit and monthly plan prices checked September 2, 2026. - Otterly pricing (https://otterly.ai/pricing). Official prompt, engine, add-on, and monthly plan prices checked September 2, 2026. - BrightLocal pricing (https://www.brightlocal.com/pricing/) and Local AI Visibility access (https://help.brightlocal.com/hc/en-us/articles/37989673412114-What-is-Local-AI-Visibility). Official active-location and plan prices checked September 2, 2026. - Semrush AI Visibility pricing (https://www.semrush.com/pricing/ai/). Official domain, prompt, engine, and price details checked September 2, 2026. - Profound pricing (https://www.tryprofound.com/pricing). Official engine, prompt, and plan prices checked September 2, 2026. - Ahrefs Brand Radar (https://help.ahrefs.com/en/articles/11064852-what-is-brand-radar-and-how-to-use-it). Official index and prompt-check prices checked September 2, 2026. Platform list covers AI Overviews and AI Mode, ChatGPT, Perplexity, Gemini, Copilot, Grok, and Claude. - Peec AI pricing (https://peec.ai/pricing). Official plan limits and model coverage checked September 2, 2026; no dollar amount was exposed in the reviewed page. - AI visibility statistics (https://www.cheers.tech/research/ai-visibility-statistics) and methodology (https://www.cheers.tech/research/methodology). Cheers Research, September 2, 2026. Per-engine citation rates across 119 home-services organizations, 28 days ending September 2, 2026. *Dylan Allen-Arnegard (https://www.linkedin.com/in/dylan-allen/) is the CEO and Co-Founder of Cheers. I help multi-location service brands get recommended by AI search and Google, with results tracked by location, employee, and competitor.* --- ## Best Local SEO Agency Alternatives for Home Services Slug: local-seo-agency-alternative-multi-location-home-services Category: Comparisons Published: 2026-06-12 Last updated: 2026-09-02 URL: https://www.cheers.tech/geo-academy/local-seo-agency-alternative-multi-location-home-services Google's own guidance on hiring SEO help, last updated June 5, 2026, now lists "Optimizing for generative AI" among the services agencies provide. Every shortlist below sells that line. The question a multi-location operator has to answer is which of them will still be there on a Tuesday when one branch's profile drifts and nobody has noticed. Last verified: September 2, 2026. The best local SEO agency alternative depends on the work that is actually failing. Scorpion, 1SEO, CAMP Digital, and RYNO are broad home-service marketing partners built around websites, SEO, paid media, lead flow, and reporting. Cheers is narrower: it manages the reviews, local pages, listings, structured facts, AI citations, and branch-level work that help each location get found and recommended. That makes the comparison more useful than calling every option an SEO agency. A brand that needs advertising and lead generation should evaluate a full-service agency. A brand that already has demand but lacks review velocity, local proof, AI cited-source tracking, and an owner for every market should evaluate Cheers. Channel eligibility still needs independent review. Apple's current Maps advertising policy prohibits home-services ads, so a full-service proposal should not count that inventory as available without a documented policy change. The Apple Maps ads guide for home services (https://www.cheers.tech/geo-academy/apple-maps-ads-home-services) covers the current boundary. Google still matters to both models. Local ranking is shaped by relevance, distance, and prominence, and complete business information helps Google match a business to relevant searches. On the review half, Google's local ranking page is unambiguous: "More reviews and positive ratings can help your business's local ranking." Google's AI Search guidance also points back to normal Search foundations: useful content, crawlable pages, good user experience, accurate business details, and structured data that matches visible content. Important: The best alternative to a local SEO agency is not "software instead of services." For a multi-location service brand, the better model is a done-for-you operating loop that connects reviews, locations, sources, and frontline behavior. If you want to pressure-test that model against one market or acquisition cluster, use the demo action below with the actual locations, services, and prompts you care about. ## The short answer Hire an agency when the work is a defined SEO project: site migration, technical cleanup, local page architecture, content planning, analytics setup, or a campaign that needs a specialist. Use software when your team already has the operators to turn alerts into work. Software can show listings gaps, rank changes, review trends, or AI visibility misses. Software alone does not guarantee that field teams, content owners, and regional managers will complete the next fix. Use a done-for-you local visibility platform when the work has to happen every week across locations. For a PE-backed HVAC group, plumbing brand, restoration rollup, garage door company, or franchise service system, the real question is not "who can write a page?" It is "who will notice the market is weak, identify the source gap, assign the work, and verify the next run?" Scorpion announced the 1SEO acquisition on June 18, 2026: "Scorpion, the leading provider of digital marketing and technology solutions for local businesses, today announced it has acquired 1SEO Digital Agency, a Philadelphia-based digital marketing agency founded in 2009." The table keeps separate rows because buyers may still encounter both brands and service models during the transition, but a current proposal should clarify which Scorpion platform, contract, and team will own the account. This is not a claim that the agencies are bad at the jobs they sell. Their public offers cover websites, paid media, lead generation, and revenue reporting. The distinction is operating focus. Cheers belongs on the shortlist when the hard problem is creating policy-compliant review opportunities, attributing them to employees and branches, tracking cited sources, and assigning local fixes every week. A full-service agency belongs on the shortlist when the buyer needs a broader acquisition program that includes advertising. For buyers comparing that model with a broad reputation platform, Birdeye Alternatives for AI Visibility and Local Reviews (https://www.cheers.tech/geo-academy/birdeye-alternatives-ai-visibility-local-reviews) lays out the tradeoff. That is where Cheers' multi-location local SEO platform (https://www.cheers.tech/solutions/multi-location-local-seo) should be evaluated: not as a generic SEO retainer, but as a system for local visibility work that depends on operations. ## The owned site is where the citations land Cheers runs a market baseline that is not a customer: 32 natural-language home-services prompts executed in 100 metro service areas across four engines, producing 356,019 citations in the 28 days ending September 2, 2026. Pooled across those engines, 68.1% of citations went to business websites and 23.0% to directory and review platforms. Google's own properties took 3.3%, social and forums 2.8%. That distribution is the strongest argument for the owned-asset work an agency and a done-for-you platform both sell. It is also why a report that only tracks rank positions misses most of what the engines are actually reading. Method and cell-size rules are in the Cheers Research methodology (https://www.cheers.tech/research/methodology), with the full cuts in AI visibility statistics (https://www.cheers.tech/research/ai-visibility-statistics). ## Where agencies are still useful A good agency can be the right choice when the problem is outside the day-to-day operating system. If a home services group just acquired 30 branches and needs URL migration support, a technical SEO agency can help. If the site is not crawlable, analytics are broken, or location pages are thin, a specialist can fix real problems. If leadership needs a content strategy, structured data review, or a second opinion on site architecture, agency work can be useful. Google's own SEO hiring guidance frames the decision this way: an SEO can improve a site and save time, but a bad one can damage the site or reputation. That is a buyer diligence problem. Ask what the agency will change, how the team will measure it, and who owns the work after the recommendation is written. For home services, the handoff is the risk. A deck that says "increase reviews in underperforming markets" is not the same as a field workflow that gets technicians, dispatchers, managers, and customers moving in the same direction. ## How to compare home-service SEO agencies If the shortlist includes Scorpion, 1SEO, CAMP Digital, RYNO Strategic Solutions, or a local home-services shop, compare the operating model behind each case-study headline. The breadth is real and it is stated plainly. RYNO's services page says: "We provide full-suite digital marketing, including SEO, website design, PPC, Local Services advertising, social media, email marketing, video production, and more." CAMP Digital's HVAC page leads with "Stay visible as AI-driven search behavior rapidly evolves." Breadth is the offer. Depth per branch is the thing to test. Ask each agency to show how it would handle five things: review velocity by branch, Google Business Profile service accuracy, location-page proof, source tracking for AI answers, and the handoff between marketing recommendations and field execution. A strong agency can still be the right partner. The gap appears when the agency can diagnose the issue but cannot move the branch behavior, review flow, or source cleanup that would fix it. The best agency relationship is honest about ownership. If the agency owns technical cleanup and page strategy while Cheers owns review generation, AI visibility tracking, and recurring local proof, the buyer gets clearer accountability than a single vague retainer promising "local SEO." ## Where retainers often need help Some traditional local SEO retainers need help in the same places: - They can report on rankings, but not always on whether the right branch appeared in ChatGPT, Gemini, Perplexity, or Google AI for a priority service query. - They can recommend more reviews, but not always attribute review creation to employees, branches, and market adoption. - They can publish service pages, but not always confirm whether the local operation can actually perform the work the page describes. - They can clean citations, but not always detect which cited sources an AI answer used this month. - They can advise profile hygiene, but not always keep the source stack aligned after acquisitions, rebrands, emergency-hour changes, and service-line changes. That does not make agencies bad. It means the buyer should separate advice from execution. A multi-location service business often needs both, but the scarce part is usually execution. Sierra Air Conditioning & Plumbing is a useful public example of the operating layer. The Sierra case study (https://www.cheers.tech/proof/sierra-cooling) ties public Google review growth to dated proof and AI visibility checks. Elite Rooter shows the plumbing version: 12 active markets and AI visibility checks across market-prompt pairs (https://www.cheers.tech/proof/elite-rooter). Those are not generic ranking reports. They are operating systems around local trust. ## What the done-for-you model should own A serious alternative to an SEO agency should own the weekly local visibility loop: - Measure priority prompts and locations across AI and local search surfaces. - Capture which sources are cited and which competitors are mentioned. - Tie review generation to branches, employees, and service moments. - Turn customer language and field proof into better location and service pages. - Keep Business Profile, location pages, citations, reviews, and structured facts aligned. - Assign every market miss to a real owner and retest after the fix. This is where AI visibility tracking (https://www.cheers.tech/solutions/ai-visibility-platform) and review generation (https://www.cheers.tech/solutions/review-generation) belong in the same conversation. AI visibility without frontline review work turns into a monitoring dashboard. Review generation without source tracking turns into a bigger review count that may not answer the buyer's next question. ## How to decide what to buy Start with the failure mode. If the site is technically broken, hire technical help. If the location pages are thin, fix the page system. If profiles and citations are inconsistent, clean the source layer. If reviews are stale, build a field workflow. If AI answers mention competitors because they cite stronger third-party sources, track the source pattern before writing another blog post. Then ask who can own the work every week. Agencies, internal teams, and software can all be part of the answer, but they should not hide the real owner. A local visibility program needs a cadence: inspect prompts and markets, identify the cited source gap, make the page or profile change, improve review capture, and check again. If the decision is moving from retainer advice into tooling, use Best Local SEO Software for Multi-Location Service Businesses (https://www.cheers.tech/geo-academy/best-local-seo-software-multi-location-service-businesses) as the comparison layer. For most multi-location service brands, the agency alternative is not cheaper content. It is tighter ownership. The right platform should make the work visible enough that marketing, operations, and local managers can act without waiting for the next retainer meeting. ## Sources - Do you need an SEO? (https://developers.google.com/search/docs/fundamentals/do-i-need-seo). Google Search Central, last updated June 5, 2026. The service list, now including optimizing for generative AI, and the warning about aggressive or spam-policy-violating work. Checked September 2, 2026. - Optimizing your website for generative AI features (https://developers.google.com/search/docs/fundamentals/ai-optimization-guide). Google Search Central, last updated July 2026. Checked September 2, 2026. - Tips to improve your local ranking on Google (https://support.google.com/business/answer/7091). Google Business Profile Help, undated help page. Relevance, distance, prominence, and the review sentence. Checked September 2, 2026. - Tips to get more reviews (https://support.google.com/business/answer/3474122). Google Business Profile Help, undated help page. Checked September 2, 2026. - Scorpion acquires 1SEO Digital Agency (https://www.scorpion.co/about-us/scorpion-news/scorpion-news/2026/scorpion-acquires-1seo-digital-agency-bringing-a/). Scorpion, June 18, 2026. Checked September 2, 2026. - Scorpion home services (https://www.scorpion.co/home-services/) and Scorpion RevenueMAX for home services (https://www.scorpion.co/revenuemax/why-revenuemax-home-services/). Scorpion, 2026. Checked September 2, 2026. - 1SEO home services digital marketing (https://1seo.com/digital-marketing-services/home-services-digital-marketing/). 1SEO, 2026. Checked September 2, 2026. - CAMP Digital HVAC marketing (https://campdigital.com/hvac-digital-marketing/). CAMP Digital, undated. AI and search optimization, paid search, websites, capacity marketing, and attribution. Checked September 2, 2026. - RYNO digital marketing services (https://rynoss.com/services/) and RYNO restoration marketing (https://rynoss.com/industries/restoration/). RYNO Strategic Solutions, undated. Full-suite service mix. Checked September 2, 2026. - AI visibility statistics (https://www.cheers.tech/research/ai-visibility-statistics) and methodology (https://www.cheers.tech/research/methodology). Cheers Research, September 2, 2026. Market Baseline citation pull: 32 prompts, 100 metros, four engines, 356,019 citations in the 28 days ending September 2, 2026. *Dylan Allen-Arnegård (https://www.linkedin.com/in/dylan-allen/) is the CEO & Co-Founder of Cheers, the done-for-you platform that manages the website, reviews, listings, structured data, and local content that get service businesses recommended across Google, Maps, ChatGPT, and Perplexity.* --- ## Why Does AI Recommend My Competitor? Slug: ai-recommends-competitors Category: Playbooks Published: 2026-06-27 Last updated: 2026-09-02 URL: https://www.cheers.tech/geo-academy/ai-recommends-competitors AI recommends a competitor when the retrieved evidence makes that business a clearer fit for the location and service in the prompt. I diagnose the miss in this order: location match, service proof, business facts, review evidence, cited sources, and engine access. Last verified: September 2, 2026. The comparison has to use the exact answer. A generic brand search cannot explain why one branch lost an emergency plumbing recommendation in one city. ## Save the evidence before changing anything Record the prompt, engine, location context, businesses named, wording of the recommendation, and every cited page. The citation trail is the only part of the mechanism the vendors actually document. OpenAI's web search guide puts it plainly: Web search allows models to access up-to-date information from the internet and provide answers with sourced citations. The same document explains how geography enters the request: "you can specify an approximate user location using country, city, region, and/or timezone." Google describes query fan-out in AI Mode, where the model issues concurrent related queries and shows supporting links. Perplexity's own developer docs describe real-time, ranked web results with citations on the answer side. None of them publish a fixed local ranking formula, and no third-party tool has one either. Google's guidance is direct about that: "No third-party tool has access to our internal ranking or AI systems." So work from the citation trail, not from a score. ## Compare the six likely gaps - Location match: does the winning source prove the competitor serves the requested market while your branch page stays vague? - Service proof: does the competitor have a clear page and recent customer language for the exact job? - Business facts: do your website, Business Profile, phone, hours, service area, and public listings agree? - Review evidence: do recent reviews describe the service and outcome in customers' own words? - Cited sources: is the engine relying on a directory, publisher, or business page where the competitor has stronger or more current evidence? - Engine access: can the engine's search crawler reach the page you expect it to use? Do not manufacture reviews, copy the competitor, or publish thin pages for every wording variation. Google says creating many pages mainly to target query variations or fan-out searches can violate its scaled content abuse policy. ## Use the panel as context The mean appearance rate is 27.0 percent in the Cheers panel, 28 days ending September 2, 2026, 119 home-services organizations, aggregate only. The median is 27.4 percent. Missing from one answer is common; missing across the high-intent prompt set is the problem to fix. The spread is what makes the average useless on its own. The 25th percentile organization appeared in 15.1 percent of its runs, the 75th in 37.3 percent, and the 90th in 47.4 percent. Only 0.8 percent of tracked organizations appeared in nothing at all, so "AI never mentions us" is almost never literally true. The home-services AI visibility index (https://www.cheers.tech/research/home-services-ai-visibility-index-2026) publishes the distribution; the methodology (https://www.cheers.tech/research/methodology) defines a run and an appearance. Each organization in that panel tracks roughly 31 buying prompts, and I only report the aggregate. Underneath it I keep the branch, prompt, engine, competitor, and source attached to every result so an average does not hide a local miss. ## Turn one miss into one owned fix If the cited page has the wrong hours, correct the hours. If the competitor wins on a specific service page, improve the real customer information and proof on the corresponding branch page. If the review set never mentions the service, fix the neutral post-service request process instead of scripting customer language. The location audit workflow (https://www.cheers.tech/geo-academy/how-to-audit-ai-search-visibility-across-locations) shows how to assign the work. For trade-specific versions of the same diagnosis, see HVAC (https://www.cheers.tech/geo-academy/how-hvac-companies-get-recommended-by-chatgpt) and plumbing (https://www.cheers.tech/geo-academy/how-plumbing-companies-get-recommended-by-chatgpt). The engine source comparison (https://www.cheers.tech/geo-academy/ai-search-engine-source-differences) explains why the winning source can change by product. Use the AI Visibility Grader (https://www.cheers.tech/visibility) for a current baseline. Keep the answer and source trail so the next run can be compared with the same scope. If the next question is whether to hand this to an agency or run it in-house, Cheers compared with an SEO agency (https://www.cheers.tech/compare/cheers-vs-seo-agency) sets out who owns which part of the work. ## Sources Every link below was opened and checked on September 2, 2026. - Web search. OpenAI Platform documentation. Checked September 2, 2026. developers.openai.com/api/docs/guides/tools-web-search (https://developers.openai.com/api/docs/guides/tools-web-search). Source of the quoted lines on sourced citations and approximate user location. - Google's guide to optimizing for generative AI features on Google Search. Google Search Central. Last updated July 10, 2026, checked September 2, 2026. developers.google.com/search/docs/fundamentals/ai-optimization-guide (https://developers.google.com/search/docs/fundamentals/ai-optimization-guide). Source of the quoted line on third-party tools, plus the query fan-out and scaled-content guidance. - Search API quickstart. Perplexity developer documentation. Checked September 2, 2026. docs.perplexity.ai/docs/search/quickstart (https://docs.perplexity.ai/docs/search/quickstart). Supports the real-time ranked web results and citation behavior described here. - Spam policies for Google web search. Google Search Central. Checked September 2, 2026. developers.google.com/search/docs/essentials/spam-policies (https://developers.google.com/search/docs/essentials/spam-policies). Supports the scaled content abuse warning against publishing a page per query variation. - Home Services AI Visibility Index 2026. Cheers Research. Cheers panel, 28 days ending September 2, 2026, 119 home-services organizations, aggregate only, about 31 tracked buying prompts per organization. Figures pulled September 3, 2026. cheers.tech/research/home-services-ai-visibility-index-2026 (https://www.cheers.tech/research/home-services-ai-visibility-index-2026). Source of the mean, median, percentile, and zero-appearance figures. *Last verified: September 2, 2026.* *Dylan Allen-Arnegard (https://www.linkedin.com/in/dylan-allen/) is the CEO and Co-Founder of Cheers. I help multi-location service brands get recommended by AI search and Google, with results tracked by location, employee, and competitor.* --- ## What Sources Does ChatGPT Use to Give Recommendations? Slug: what-sources-does-chatgpt-use Category: Research Published: 2025-12-20 Last updated: 2026-09-02 URL: https://www.cheers.tech/geo-academy/what-sources-does-chatgpt-use Across the 28 days ending September 2, 2026, the Cheers market baseline collected 70,723 ChatGPT citations from home-services prompts run in 100 metro areas. 48.8% of them pointed at a contractor's own website. 44.7% pointed at a directory or review platform such as Angi, BBB, Yelp, or Expertise. Social and forum sources took 4.3%, media and vendor pages 2.0%, and google.com 0.2%. So the practical answer is two sources, roughly split: your website and the directories that describe you. Everything else is rounding. Last verified: September 2, 2026. OpenAI does not publish a retrieval map, and its help article on searching the web is explicit that nothing is owed to any site: ChatGPT ranks search results using multiple factors intended to help users find relevant, reliable information. Placement is not guaranteed. The full panel cuts, including the same split for Gemini, Perplexity, and Google AI Overviews, are in AI visibility statistics (https://www.cheers.tech/research/ai-visibility-statistics), with the sampling rules in the methodology (https://www.cheers.tech/research/methodology). To see which sources come back for your own brand and markets, run the Cheers AI Visibility Grader (https://www.cheers.tech/visibility). ## The short answer Visible citations and published studies show local-business answers drawing from review platforms, directories, websites, and news articles. The internal source mix varies by product, mode, prompt, and date, and no public source provides a complete ranking or retrieval map. For the broader strategic frame, read What is Generative Engine Optimization? (https://www.cheers.tech/geo-academy/what-is-geo). For Google's own AI Search advice, read How Local Businesses Can Show Up in Google AI Search (https://www.cheers.tech/geo-academy/how-local-businesses-can-show-up-in-google-ai-search). If ChatGPT is already showing the wrong hours, phone, service, or branch, use the business-information correction workflow (https://www.cheers.tech/geo-academy/fix-wrong-business-information-chatgpt) to trace the visible source and verify the fix by market. ## The longer answer Modern AI assistants can combine model knowledge with web search or retrieval features. The balance is product- and query-specific. Model knowledge can contain information learned before the current query, but vendors do not publish a complete business-by-business inventory of that knowledge. Search and retrieval features can fetch current web information when the product and query use them. A local prompt does not guarantee that every available source will be searched or cited. Machine-readable business information includes structured data on business websites and platform-managed business records. Public documentation does not establish one universal weighting for those inputs. OpenAI's help article, updated in late August 2026, describes the retrieval step in plain terms. ChatGPT rewrites your question into one or more targeted queries, sends those to search providers, and shares an approximate location taken from your IP address so that a "near me" question becomes a city-scoped query. OpenAI gives site owners exactly one concrete eligibility instruction: allow OAI-SearchBot to crawl the site, and make sure the host or CDN is not blocking OpenAI's published searchbot IP ranges. That is a robots and firewall check, not a content tactic. Tip: Record the pages visibly cited for the exact prompt you tested. Use those pages as an audit trail, not as a permanent map of the engine's ranking system. ## Sources that matter for local businesses Google Business Profile supplies business information to Google's local systems, and Google's own guidelines warn that duplicate profiles cause display problems on Maps and Search. Whether ChatGPT reads that record directly is unestablished. In May 2025, Search Engine Land's Damian Rollison described ChatGPT running a Bing search for the local query, taking 20 to 30 of the top results, and reordering them with its own logic, while not reading Bing Places profile data directly. Local Falcon's November 2025 teardown observed the same shape and named the platforms it saw most: Yelp, TripAdvisor, Better Business Bureau, MapQuest, and Yellow Pages. Both are dated observations of a system that changes, and OpenAI's current documentation names no provider except in the restaurant-reservation case, where availability may come from OpenTable, Resy, or Yelp. The Cheers baseline gives the same picture from the citation side. The most cited platform domains across all 100 metros, all four engines, were google.com, expertise.com, angi.com, bbb.org, and reddit.com. Every one of those has a record about your business that you can read today. Review platforms like Yelp, Facebook, TripAdvisor, and industry-specific sites can appear in search results and citations. They provide customer-authored context, but the products do not publish a universal sentiment or review-ranking formula. Your website can be retrieved when it is crawlable and relevant to the query. Clear service descriptions and accurate schema help machines interpret the page, but they do not guarantee a citation or recommendation. Directory listings can supply business facts and appear in retrieved results. Keep relevant listings accurate, especially phone numbers, addresses, hours, URLs, and services. News and press coverage may appear when it is relevant to the query. Earned coverage should be factual and useful, not created to simulate authority. For a local operator, conflicting public facts create a customer problem and a source-audit target. If Bing-visible pages, Yelp, BBB, Apple Maps, and the company website disagree on service area or phone number, correct the sources customers and crawlers can reach. Do not claim that consistency alone causes a recommendation. For the plumbing-specific version of this source map, read How Do Plumbing Companies Get Recommended by ChatGPT? (https://www.cheers.tech/geo-academy/how-plumbing-companies-get-recommended-by-chatgpt). ## What this means for your strategy You can't control which sources a particular AI query uses. But you can ensure your business is well-represented across all the major sources. Important: Maintain your Google Business Profile, then inspect the other sources that actually appear for your markets and prompts. Do not create or maintain a listing only because it appears in a generic platform checklist. Earn reviews where customers use the platform. Keep relevant profiles accurate and follow each platform's solicitation policy. Do not manufacture a multi-platform review footprint for an assumed AI benefit. Implement applicable schema markup. It should describe the visible page accurately and should not be treated as a recommendation guarantee. Maintain business-information accuracy. Correct wrong or stale names, addresses, phone numbers, hours, URLs, and services. Harmless formatting differences are not the same as conflicting facts. Keep information current. Update profiles and pages when hours, services, locations, contact paths, or policies change. Old data is not always worse than no data, but incorrect data can harm the customer journey. Build a source stack. Your website should explain services and locations clearly. Your listings should match the same facts. Your reviews should show recent customer experience. Your schema should mark up the business in a machine-readable way. If you need the technical piece, see What Is JSON-LD? (https://www.cheers.tech/geo-academy/what-is-json-ld). If you need the review piece, see How Reviews Support AI Visibility for Local Businesses (https://www.cheers.tech/geo-academy/reviews-and-geo). Important: You cannot control which source a query retrieves. You can keep the public sources you own or manage accurate, useful, and consistent with the real business. ## Sources Every link below was opened and checked on September 2, 2026. - Searching the web with ChatGPT (https://help.openai.com/en/articles/9237897). OpenAI Help Center, updated August 2026. Query rewriting, search partners, approximate location, reservation providers, ranking, and OAI-SearchBot eligibility. - How does ChatGPT conduct local searches? (https://searchengineland.com/how-does-chatgpt-conduct-local-searches-454894). Search Engine Land, May 2025, by Damian Rollison. The Bing retrieval step and the reordering that follows. - ChatGPT local search data sources (https://www.localfalcon.com/blog/chatgpt-local-search-data-sources-where-does-business-info-come-from). Local Falcon, November 2025. The review and directory platforms observed most often. - Guidelines for representing your business on Google (https://support.google.com/business/answer/3038177). Google Business Profile Help. One profile per business, and what duplicates do to Maps and Search. - Optimizing your website for generative AI features on Google Search (https://developers.google.com/search/docs/fundamentals/ai-optimization-guide). Google Search Central, last updated July 2026. - AI Visibility in 2025: How Gemini, ChatGPT, and Perplexity Cite Brands (https://www.yext.com/blog/ai-visibility-in-2025-how-gemini-chatgpt-perplexity-cite-brands). Yext, October 2025. 6.8 million citations across 1.6 million responses. - Cheers market baseline, 28 days ending September 2, 2026 (https://www.cheers.tech/research/ai-visibility-statistics). 32 home-services prompts, 100 metro areas, four engines, 70,723 ChatGPT citations, aggregate only. Methodology (https://www.cheers.tech/research/methodology). *Dylan Allen-Arnegård (https://www.linkedin.com/in/dylan-allen/) is the CEO of Cheers, the local search platform for service businesses.* --- ## What Is Query Fan-Out in Google AI Mode? Slug: what-is-query-fan-out-google-ai-mode Category: AI Search News Published: 2026-05-31 Last updated: 2026-09-02 URL: https://www.cheers.tech/geo-academy/what-is-query-fan-out-google-ai-mode Google defines query fan-out in its own optimization guide: A set of concurrent, related queries generated by the model to request more information and fetch additional relevant search results to address the user's query. Last verified: September 2, 2026. Google's worked example starts with "how to fix a lawn that's full of weeds" and fans out into things like best herbicides for lawns and removing weeds without chemicals. A local service request behaves the same way: one question about a failed water heater turns into several searches about the job, the market, the proof, and who is available. Google does not publish the exact hidden set for any given answer. The panel data shows the practical consequence. Across the Cheers panel of 119 home-services organizations, 28 days ending September 2, 2026, aggregate only, the shape of the question changed the outcome more than anything an operator does on the page that week. Price and cost questions returned a 37.9% pooled appearance rate. Broad "best or top" prompts without a place name returned 28.9%. Question-shaped prompts, the how and what and should-I family, returned 26.4% appearance but the highest citation rate of any family at 27.3%. "Best near me" and named-place prompts came last at 22.8%. That is the argument for writing pages that answer the whole buying decision instead of chasing one phrase. The cuts are in AI visibility statistics (https://www.cheers.tech/research/ai-visibility-statistics) and the sampling rules in the methodology (https://www.cheers.tech/research/methodology). ## What Google confirms Google says AI Mode and AI Overviews use core Search ranking and quality systems. Retrieval-augmented generation pulls current pages from the Search index. Query fan-out generates related searches. The system then shows supporting links for the response. That means the technical foundation still matters. A useful page has to be crawlable, indexed, eligible for a Search snippet, and linked clearly from the rest of the site. Google says there is no special AI schema or extra technical requirement. ## What local operators should do Write for the buying decision, not an imagined list of subqueries. A strong location page should state the services that branch truly provides, the area it serves, current contact and booking details, hours or availability, qualifications, and local proof. The location-page checklist (https://www.cheers.tech/geo-academy/what-should-location-pages-include-for-ai-search) covers that structure. Service pages should answer the parts of the job that change the buyer's choice, such as what is included, who performs the work, how estimates work, and what evidence supports the claim. Do not produce a thin page for every city-service wording. Google names the tactic and the policy directly, and then explains why it fails on its own terms: a high quantity of pages doesn't make a website higher quality or more relevant to users Creating separate content for every search variation or fan-out query, primarily to manipulate rankings or generative AI responses, violates Google's scaled content abuse spam policy. The safer version of that instinct is building city pages without doorway risk (https://www.cheers.tech/geo-academy/build-city-pages-without-doorway-risk-ai-search). ## How I audit a result I save the original prompt, AI Mode response, supporting links, market, and time checked. Then I compare the pages Google cited with the branch page and public business facts. If a competitor page answers a real buyer concern more clearly, improve the useful content. If a source contains the wrong location or hours, correct the source. Do not claim to have reverse-engineered the hidden query tree from one response. The AI engine source comparison (https://www.cheers.tech/geo-academy/ai-search-engine-source-differences) explains why AI Mode should be tracked separately from ChatGPT, Perplexity, and Copilot. Does SEO Still Work? (https://www.cheers.tech/geo-academy/does-seo-still-work) covers why the core Search work still matters. Use the AI Visibility Grader (https://www.cheers.tech/visibility) when you need a current baseline for your own business. ## Sources Every link below was opened and checked on September 2, 2026. - Optimizing your website for generative AI features on Google Search (https://developers.google.com/search/docs/fundamentals/ai-optimization-guide). Google Search Central, last updated July 2026. Source of the query fan-out definition, the lawn-weeds example, the scaled content abuse warning, and the note that structured data is not required for generative AI search. - AI features and your website (https://developers.google.com/search/docs/appearance/ai-features). Google Search Central. Eligibility, indexing, internal-link, and Search Console guidance for AI surfaces. - Scaled content abuse (https://developers.google.com/search/docs/essentials/spam-policies). Google Search Essentials spam policies. The policy Google names when it warns against a page per query variation. - Cheers panel, 28 days ending September 2, 2026 (https://www.cheers.tech/research/ai-visibility-statistics). 119 home-services organizations, pooled appearance and citation rates by prompt shape, aggregate only. Methodology (https://www.cheers.tech/research/methodology). *Dylan Allen-Arnegard (https://www.linkedin.com/in/dylan-allen/) is the CEO and Co-Founder of Cheers. I help multi-location service brands get recommended by AI search and Google, with results tracked by location, employee, and competitor.* --- ## What Should Location Pages Include for AI Search? Slug: what-should-location-pages-include-for-ai-search Category: Playbooks Published: 2026-05-25 Last updated: 2026-09-02 URL: https://www.cheers.tech/geo-academy/what-should-location-pages-include-for-ai-search A location page should include the branch name, address or service-area model, phone, hours, services, coverage, qualifications, local proof, reviews, staff context, and one direct contact or booking path. The structured data should match those visible facts. Last verified: September 2, 2026. ## The branch checklist - Identity: the public-facing branch name and the brand relationship. - Location model: a customer-facing address or a clear service-area explanation. - Contact path: the correct local phone and one direct form or booking action. - Hours and availability: current operating, emergency, or appointment details. - Services: the work that branch actually provides, with useful links to service detail. - Coverage: cities or areas the branch truly serves, without stuffing place names. - Qualifications: licenses, certifications, warranties, or insurance claims that are current and verifiable. - Local proof: project photos, staff, community context, and customer evidence tied to that market. - Reviews: real, relevant customer feedback presented within platform and schema policies. - Structured data: LocalBusiness details that match the visible page and canonical branch facts. ## What our own page audits showed Between May 12 and May 19, 2026 we audited 371 location pages across 6 organizations. Every one returned a successful status code, and the average page score was 82.0 with 2.4 detected issues, yet only 41 of those pages published LocalBusiness structured data. The pages loaded fine; the machine-readable branch facts were missing. That is a narrow anonymized product sample, not a market benchmark. It does match what I keep running into on branch pages: the page reads well to a customer and still leaves a crawler unable to say which business it describes. ## Why one parent page is not enough A buyer choosing a service company needs to know whether the right team serves the market and performs the job. A parent-brand page can explain the company, but it often cannot answer branch hours, service coverage, phone routing, or local proof. Google's business-link policy leaves no room here: For businesses with multiple locations, action links must lead to a website for a specific location. The same policy adds that local business links must allow customers to complete the designated action, which rules out sending a booking link to a page that only lists a phone number. Google's AI Search guidance points the same direction: keep Business Profile information current and make important content available as text. ## Keep every fact aligned The page, Business Profile, listings, call routing, and schema should describe the same real operation. Fix material conflicts such as the wrong branch phone, old address, closed hours, or unsupported service. Do not turn small formatting differences into an emergency. The goal is a correct customer journey and a source trail an engine can retrieve without finding two different businesses. The citation cleanup guide (https://www.cheers.tech/geo-academy/citations-stack-2025) covers those conflicts. The duplicate Business Profile guide (https://www.cheers.tech/geo-academy/fix-duplicate-google-business-profiles-ai-search) covers multiple records for the same operation. ## Write for a customer, not a city template Google warns against scaled pages created mainly to manipulate Search or generative responses. Its generative AI guidance also closes off the shortcut people ask about most: You don't need to create new machine readable files, AI text files, markup, or Markdown to appear in Google Search. A useful branch page earns its URL by carrying real local information and proof, not by publishing a parallel file for machines. Our panel puts numbers on what branch pages are competing for: across 119 home-services organizations in the 28 days ending September 2, 2026, prompts naming a place or asking for something near me returned a pooled appearance rate of 22.8%, the lowest of any published prompt shape in the home services AI visibility index (https://www.cheers.tech/research/home-services-ai-visibility-index-2026). I would rather publish one complete branch page than many thin service-area combinations. Use service pages for meaningful service detail and link them clearly from the branch page. Use the AI Visibility Grader (https://www.cheers.tech/visibility) to see whether the public location evidence is reaching current AI answers. Save the prompt and cited sources before deciding what the page needs next. ## Sources Checked September 2, 2026. - Business links policies and guidelines. Google Business Profile Help, accessed September 2026. support.google.com/business/answer/13769188 (https://support.google.com/business/answer/13769188?hl=en). Source of the quoted location-specific action-link rule and the complete-the-action requirement. - Optimizing your website for generative AI features on Google Search. Google Search Central, last updated July 10, 2026. developers.google.com/search/docs/fundamentals/ai-optimization-guide (https://developers.google.com/search/docs/fundamentals/ai-optimization-guide). Source of the quoted machine-readable-files sentence and the scaled-content guidance. - Local business (LocalBusiness) structured data. Google Search Central, last updated December 10, 2025. developers.google.com/search/docs/appearance/structured-data/local-business (https://developers.google.com/search/docs/appearance/structured-data/local-business). Source of the LocalBusiness markup guidance. - Cheers website optimization audits, May 12 to 19, 2026: 371 location pages across 6 organizations, aggregate only. Source of the structured-data detection counts. - Cheers panel: 119 home-services organizations, 28 days ending September 2, 2026, aggregate only. Home services AI visibility index (https://www.cheers.tech/research/home-services-ai-visibility-index-2026) and methodology (https://www.cheers.tech/research/methodology). Source of the pooled appearance rate for place-named prompts. *Dylan Allen-Arnegard (https://www.linkedin.com/in/dylan-allen/) is the CEO and Co-Founder of Cheers. I help multi-location service brands get recommended by AI search and Google, with results tracked by location, employee, and competitor.* --- ## How Reviews Support AI Visibility for Local Businesses Slug: reviews-and-geo Category: Reviews Published: 2025-10-05 Last updated: 2026-09-02 URL: https://www.cheers.tech/geo-academy/reviews-and-geo BrightLocal's 2026 Local Consumer Review Survey, fielded with 1,002 US adults, reports that 97% of consumers read reviews for local businesses and that 74% seek reviews written in the last three months. Recency is now part of what a review profile has to prove, and that is a measurable operating target rather than a ranking theory. Last verified: September 2, 2026. The survey puts the recency demand plainly: "74% seek reviews written in the last three months." Google says review count and score can factor into local prominence. For AI products the public evidence is thinner: review pages may appear in retrieved or cited results, but no provider publishes a universal review-ranking formula. So the honest frame is evidence, not weighting. Important: Treat reviews as customer evidence, not as a keyword feed. Ask neutrally, never script the customer's words, and do not claim that a phrase or sentiment score causes an AI recommendation. If you are new to the broader category, start with What is Generative Engine Optimization? (https://www.cheers.tech/geo-academy/what-is-geo). This article focuses on the review layer of that system. ## What a public review can show Review pages contain text that customers can read and web retrieval systems may access. The exact treatment of that text varies by platform and query. A review that says "John was on time, explained the repair clearly, and the price was fair" gives a prospective customer more useful context than a rating alone. It documents punctuality, communication, and pricing from that customer's perspective. "Great service, five stars" is still legitimate feedback, but it gives a future reader less detail about the work. Detailed reviews can give customers more context than vague praise. Volume and detail answer different questions, so compare them separately. This is why review quality matters as much as quantity. The compliance line matters too: ask every eligible customer for honest feedback, but do not script exact words, request keywords, or ask for a specific rating. Google's own guidance on asking for reviews is comfortable with the downside: "Honest and balanced reviews can help potential customers decide. A mix of positive and negative feedback often feels more trustworthy." For the full policy guardrails, see Compliance Playbook: Collect More Reviews Without Getting Flagged (https://www.cheers.tech/geo-academy/compliance-playbook). ## Use velocity as an operating metric Businesses change over time, so customers often look for current feedback. A company with 2,000 reviews from 2019 and no recent activity gives buyers less current evidence than a profile that reflects recent work. Review velocity is the rate at which new reviews arrive. Track it to understand whether the request process is active and whether each branch has current customer feedback. Do not treat it as a published AI ranking weight. Our own panel shows how wide the operating spread is. In the Cheers production review pull covering the 90 days ending September 2, 2026, across 73 home-services organizations with at least one review, the median organization collected 6.0 new reviews a month while the 90th percentile collected 46.7. Google supplied 97.9% of those reviews, and the owner response rate across the group was 81.9%. Aggregate only, no organization named; the full methodology (https://www.cheers.tech/research/methodology) and the rest of the panel are in the Home Services AI Visibility Index (https://www.cheers.tech/research/home-services-ai-visibility-index-2026). That spread is a process gap. The 90th percentile is not asking better customers; it is asking more of them, at the moment the work finishes. Tip: Compare lifetime count and recent activity separately. They answer different operating questions, and neither guarantees visibility. In a visibility audit, recent reviews can show that a location is active and give customers current service context. That is useful evidence even when no causal ranking effect can be established. For example, a pest control branch with 160 lifetime reviews and 24 in the last 60 days gives customers more recent context than a profile with no recent activity. The larger, older profile may still perform better in Search or AI results; the review counts alone cannot predict the outcome. ## Responses are for customers Review responses are public and may be available to a retrieval system that accesses the page. Their primary purpose is still customer communication. BrightLocal's 2026 survey found that 89% of consumers expect business owners to respond to reviews, and 80% say they are likely to use a business that responds to all of its reviews. Google's reply guidance points the same way, telling owners to "Be professional and polite: Keep your replies clear and helpful." When you respond, address the specific issue where privacy and the known facts allow it. A copied template can frustrate customers because it does not answer their concern. Silence does not prove that a business is overwhelmed or indifferent, but unanswered complaints can leave future customers without the company's side of the story. Tip: Respond where a response helps. Verify the facts, protect private information, explain the next step, and move account-specific resolution to a private channel. ## Choose review platforms by customer relevance Google reviews are prominent in Google Search and Maps. Yelp, Facebook, BBB, and industry directories may also matter to customers or appear in visible source sets, depending on the category and query. Local Falcon's November 2025 teardown of ChatGPT's local data found that "ChatGPT leans heavily on established review and directory platforms, including: Yelp, Tripadvisor, Better Business Bureau (BBB), MapQuest, Yellow Pages". That is one vendor's observation of one engine, not a published source list, so treat it as a place to look rather than a rule. Reviews across relevant platforms give customers more places to verify a business and may appear in different products' source sets. The effect depends on the category, platform, query, and quality of the underlying profile. Important: If you have 4.8 stars on Google and 3.2 on Yelp, inspect why. The audiences, sample sizes, dates, or service mix may differ. Do not assume the discrepancy itself triggers an AI penalty. ## Review language helps customers understand the work Star ratings compress a customer's experience into one number. The written review can explain what happened, what service was performed, and why the customer chose that rating. "The job was fine, I guess" communicates something different from "The job exceeded my expectations," even when both reviews use the same star rating. That distinction is useful for service coaching and for prospective customers reading the profile. Customers must choose their own words. A neutral request can invite honest detail, but it must not steer the rating, keywords, service claims, or sentiment. Tip: Keep the request neutral: "If you'd like to share an honest review of today's service, this link will take you there." Do not request a topic, rating, employee name, or particular wording. ## Building a compliant review program The strategy isn't complicated, but it requires consistency: Use a consistent handoff. NFC badges, QR codes, and follow-up messages can all provide a direct link. Make the request neutral and optional under one eligibility rule, then let the customer complete it privately. Hello Sugar (https://www.cheers.tech/proof/hello-sugar) used NFC badges to increase review activity from 50 to 700 reviews per month; that case study describes an operating result, not an AI ranking effect. Train your team to ask. The branches with the best review velocity aren't lucky. They've built review requests into their service process. Every technician and every visit should follow the same compliant workflow. If you would rather hand that operating load to someone else, review generation (https://www.cheers.tech/solutions/review-generation) is the service line that covers it. Respond where it helps the customer. Use specific, factual responses, protect private information, and avoid arguing in public. Use How to Respond to Negative Reviews Professionally (https://www.cheers.tech/geo-academy/responding-to-negative-reviews) when the team needs a response standard for unhappy customers. Maintain relevant profiles. Focus on the platforms customers in the category actually use or that appear in the answers you audit. Do not create review profiles only to satisfy a diversity checklist. Watch recent activity and lifetime count. Monthly trends show whether the request process is active; lifetime totals show the depth of the profile. Neither one universally matters more for AI recommendations. Connect reviews to the rest of your GEO system. Reviews work best when your Google Business Profile, website, citations, and schema say the same thing. If you want to understand the source layer behind this, read What Sources Does ChatGPT Use to Give Recommendations? (https://www.cheers.tech/geo-academy/what-sources-does-chatgpt-use). If you need the field workflow, read Review Collection at Point of Service: A Playbook (https://www.cheers.tech/geo-academy/review-collection-best-practices). If Google has generated a short synthesis above a location's recent reviews, use the Google AI review summary audit (https://www.cheers.tech/geo-academy/change-google-ai-review-summary-business) to distinguish that output from editable profile copy, classify the supporting theme, and route a real operating problem to the branch that owns it. Reviews remain social proof and customer feedback. They can also appear on pages that search and AI products retrieve. Build a representative, policy-compliant review profile because it helps customers and operators, then measure AI visibility separately. ## Sources - Tips to get more reviews (https://support.google.com/business/answer/3474122). Google Business Profile Help, undated help page. Checked September 2, 2026. - Manage customer reviews (https://support.google.com/business/answer/3474050). Google Business Profile Help, undated help page. Checked September 2, 2026. - Local Consumer Review Survey 2026 (https://www.brightlocal.com/research/local-consumer-review-survey/). BrightLocal, 2026, 1,002 US adult consumers. Checked September 2, 2026. - ChatGPT Local Search Data Sources (https://www.localfalcon.com/blog/chatgpt-local-search-data-sources-where-does-business-info-come-from). Local Falcon, November 2025. Checked September 2, 2026. - New ways we are protecting businesses on Maps (https://blog.google/products-and-platforms/products/maps/new-ways-were-protecting-businesses-on-maps/). Google, April 2026. Checked September 2, 2026. - Optimizing your website for generative AI features (https://developers.google.com/search/docs/fundamentals/ai-optimization-guide). Google Search Central, last updated July 2026. Checked September 2, 2026. - Home Services AI Visibility Index 2026 (https://www.cheers.tech/research/home-services-ai-visibility-index-2026). Cheers Research, September 2, 2026. Panel methodology at /research/methodology (https://www.cheers.tech/research/methodology). To put these principles into practice with your field team, see Review Collection at Point of Service: A Playbook (https://www.cheers.tech/geo-academy/review-collection-best-practices). *Dylan Allen-Arnegård (https://www.linkedin.com/in/dylan-allen/) is the CEO of Cheers, the local search platform for service businesses.* --- ## What is Generative Engine Optimization (GEO)? Slug: what-is-geo Category: Playbooks Published: 2025-09-15 Last updated: 2026-09-02 URL: https://www.cheers.tech/geo-academy/what-is-geo Across our own panel of 119 home-services organizations, the median business was named in 27.4% of the AI checks run on its behalf during the 28 days ending September 2, 2026. The top decile hit 47.4%. The bottom quartile sat at 15.1%. Nobody in that panel gets recommended every time, and almost nobody gets recommended never. That spread is what Generative Engine Optimization is actually about. Ask ChatGPT for an HVAC company in Las Vegas, or Gemini for a waxing salon in Dallas, and the answer may name businesses, cite web pages, show local results, or decline to recommend anyone. GEO is the work of measuring those answers and improving the public evidence a business can verify. For local operators, GEO is not a hidden technical switch. It is a repeatable process: test the buyer questions that matter, record whether each location appears, inspect the visible sources, correct material errors, and retest after the source changes can be crawled. Tip: GEO adds answer-level measurement to SEO. It does not replace crawlability, useful pages, local profiles, or customer trust. For a local service operator, this is not an abstract marketing trend. It is a practical evidence problem. A 40-location HVAC brand needs every market to show the same business facts, fresh reviews, clear service pages, and proof that technicians actually solve customer problems. A single strong homepage cannot carry weak location pages. ## What GEO changes Search reporting usually starts with rankings, impressions, clicks, and local-pack visibility. GEO adds questions that those reports do not answer: Did the business appear in the generated response? Was the right branch named? Which pages were cited? Did a competitor appear instead? Were the phone number, service area, and offer accurate? An AI answer may name one business, several businesses, or none. The format varies by product and prompt, so the operator should save the exact answer rather than assuming every engine behaves like a shorter search-results page. ## What the distribution looks like A GEO baseline is only readable next to a population. Ours covers 119 home-services organizations and 2,671,846 checks in the 28 days ending September 2, 2026. The quartiles matter more than the mean. A brand at 15.1% and a brand at 47.4% are both normal in this panel, and the gap between them is roughly three times, which is large enough to change a month. Only 0.8% of these organizations never appeared at all, so the common fear that AI simply cannot see a local business is not what the data shows. What it shows is a wide, workable middle. The full distribution, plus the same cut by trade and prompt shape, is in the home services AI visibility index (https://www.cheers.tech/research/home-services-ai-visibility-index-2026), with the appearance formula in the research methodology (https://www.cheers.tech/research/methodology). ## What operators can actually inspect AI assistants do not expose one universal local-business ranking system. The reliable evidence is what the operator can observe: The answer. Save the business names, wording, provider, prompt, market, and date. The citations. Record the exact pages shown with the answer. A citation proves that a page supported that response, not that the page has a permanent ranking weight. Public business facts. Check names, locations, services, hours, phone numbers, and booking paths across the business website and relevant profiles. Owned pages. Confirm that location and service pages are crawlable, internally linked, useful to a buyer, and marked up consistently with their visible content. Customer evidence. Maintain a policy-compliant review process because reviews help customers evaluate a business. Do not claim that review count, recency, sentiment, or particular wording has a published universal AI weight. Important: Treat every suspected signal as an audit hypothesis until the answer, citation, or official product documentation supports it. For a deeper breakdown of where ChatGPT-style systems may retrieve local business facts, see What Sources Does ChatGPT Use to Give Recommendations? (https://www.cheers.tech/geo-academy/what-sources-does-chatgpt-use). For Google's current advice, see How Local Businesses Can Show Up in Google AI Search (https://www.cheers.tech/geo-academy/how-local-businesses-can-show-up-in-google-ai-search). ## GEO and SEO are different measurements SEO asks whether pages can be crawled, indexed, understood, and surfaced for relevant searches. GEO asks what happened in a defined set of generated answers and which sources were visible. A page can rank in Google without being cited in one AI answer, and an AI answer can change by provider, prompt, market, or date. Google's guidance for AI features points site owners back to normal Search fundamentals, and its reasoning is worth quoting rather than paraphrasing: The best practices for SEO continue to be relevant because our generative AI features on Google Search are rooted in our core Search ranking and quality systems. The same guide, updated July 10, 2026, adds that you do not need new machine-readable files, AI text files, markup, or Markdown to appear in Google Search. GEO work should strengthen the ordinary foundations first, then add provider-level testing and source evidence. ## Where reviews fit Reviews are public customer evidence. They can help a buyer understand recent service experiences, and review pages may appear in search results or citations. No public source establishes a universal AI formula for review volume, velocity, sentiment, or wording. Use one neutral eligibility rule, make every request optional, and follow each platform's policy. Measure review activity to improve the customer-feedback process, not to claim a guaranteed AI-ranking effect. For the evidence and policy boundaries, read How Reviews Support AI Visibility for Local Businesses (https://www.cheers.tech/geo-academy/reviews-and-geo). ## What this means for your business Start with a baseline for the services and markets that create revenue. Save the answers and citations, separate wrong facts from missing coverage, and assign each fix to the team that owns the source. Correct customer-facing errors first. Then improve thin location or service pages, relevant profiles, internal links, and applicable structured data. Retest the same prompts after the changes have had time to be crawled. If you need the technical layer, start with What Is JSON-LD? (https://www.cheers.tech/geo-academy/what-is-json-ld). If you need the review layer, read How Reviews Support AI Visibility for Local Businesses (https://www.cheers.tech/geo-academy/reviews-and-geo). You can also run the free AI Visibility Grader (https://www.cheers.tech/visibility) for a one-profile snapshot. For a more detailed operating sequence, use The GEO Playbook: Get Recommended by AI (https://www.cheers.tech/geo-academy/how-to-get-ai-to-recommend-my-business). ## Sources Checked September 2, 2026. - Optimizing your website for generative AI features on Google Search. Google Search Central, last updated July 10, 2026. developers.google.com/search/docs/fundamentals/ai-optimization-guide (https://developers.google.com/search/docs/fundamentals/ai-optimization-guide). Source of the quoted core-systems sentence and the machine-readable-files guidance. - How does ChatGPT conduct local searches? Damian Rollison, Search Engine Land, May 2, 2025. searchengineland.com (https://searchengineland.com/how-does-chatgpt-conduct-local-searches-454894). Source of the reported ChatGPT local retrieval mechanics. - Local Consumer Review Survey 2026. BrightLocal, 2026. brightlocal.com/research/local-consumer-review-survey (https://www.brightlocal.com/research/local-consumer-review-survey/). Source of the consumer review behavior context. - LocalBusiness. Schema.org vocabulary, accessed September 2026. schema.org/LocalBusiness (https://schema.org/LocalBusiness). Source of the structured data vocabulary. - Tips to get more reviews. Google Business Profile Help, accessed September 2026. support.google.com/business/answer/3474122 (https://support.google.com/business/answer/3474122?hl=en). Source of the permitted review-request guidance and the incentive prohibition. Replaces the previous link to the Business Profile help home page. - Cheers panel: 119 home-services organizations, 2,671,846 checks, 28 days ending September 2, 2026, aggregate only. Home services AI visibility index (https://www.cheers.tech/research/home-services-ai-visibility-index-2026) and methodology (https://www.cheers.tech/research/methodology). Source of the appearance-rate quartiles and the share that never appeared. *Dylan Allen-Arnegård (https://www.linkedin.com/in/dylan-allen/) is the CEO of Cheers, the local search platform for service businesses.*