To appear in ChatGPT and Perplexity recommendations, contractors need to complete their Google Business Profile, add LocalBusiness and FAQPage schema to their website, match their name and address exactly across all directories, and build a consistent flow of recent reviews. These seven steps, implemented in order, are the complete optimization sequence for AI search visibility in 2026.
of consumers now use AI tools for local service discovery
Not Google. Not Yelp. They open ChatGPT or Perplexity and ask who to call. The contractor AI recommends gets the call. The contractor not in the results doesn't exist to that customer.
How Does ChatGPT Decide Which Businesses to Recommend?
ChatGPT doesn't have a ranking algorithm you can pay into. It doesn't respond to keywords the way Google does. Instead, it synthesizes data from the sources it trusts most for local business information: your Google Business Profile, your website's structured data, your review count and recency across platforms, and your citation footprint across directories.
ChatGPT uses web crawling through Bing's index (via the browsing tool) alongside its training data, which includes structured business data from aggregators like Yelp, Angi, HomeAdvisor, and the BBB. It builds an internal confidence score for each business it encounters — a measure of how clear, complete, and consistent the signals are. The business with the highest confidence score wins the recommendation.
Perplexity works differently in one key way: it crawls the live web more aggressively and is more likely to cite specific pages as sources. Where ChatGPT relies heavily on GBP and aggregator data, Perplexity rewards contractors who have answer-ready pages — content that directly answers the questions customers type in. A page titled "How Much Does Furnace Replacement Cost in Denver?" that actually answers the question is the kind of content Perplexity surfaces and cites by name.
The good news: the same seven optimization steps improve your standing in both systems. They also improve your AEO vs. SEO performance simultaneously — the signals that AI systems use and the signals that affect Google rankings overlap more than most contractors realize. For a breakdown of what agencies charge for this work, see AEO Pricing: What Agencies Charge.
The 7-Step ChatGPT Optimization Process for Contractors
These steps are ordered by impact and implementation speed. Start with step one and work forward — each step builds on the last, and the compounding effect across all seven is significantly stronger than any single fix alone.
Complete your Google Business Profile to 100%
GBP is the single most important data source AI systems use for local contractor recommendations. Every field matters: list each service individually (not just "plumbing" — "drain cleaning," "water heater installation," "pipe repair"), confirm your service area at the zip-code level, write a business description that names the cities you serve and the problems you fix, add recent photos, and answer every question in the Q&A section. An incomplete GBP is the most common reason contractors are invisible in ChatGPT. This step takes under two hours and has immediate effect.
Add LocalBusiness schema markup to your website
Schema markup is structured data embedded in your website's code that tells AI crawlers exactly what your business does. For contractors, the highest-priority type is LocalBusiness with the correct sub-type for your trade — Plumber, HVACBusiness, ElectricalContractor, RoofingContractor, GeneralContractor. This JSON-LD block should live on your homepage and include your business name, address, phone number, service area, hours, and aggregate review rating. Without schema, AI has to infer your services from page text and often gets it wrong — or skips you entirely for a competitor whose data is explicit.
Add Service and FAQPage schema to each service page
One schema block per service page, specifying the service name, a plain-language description, and the geographic area served. FAQPage schema on any page that has a question-and-answer section tells Perplexity and Google AI Overview exactly which questions you answer and what the answers are — making your content directly quotable. This is the mechanism by which a page on your site becomes a cited source in an AI answer. Contractors with FAQ schema consistently appear in AI answers for cost, timeline, and comparison questions.
Audit and repair citation consistency across directories
Your Name, Address, and Phone number (NAP) must be formatted identically across every directory AI pulls from: Google Business Profile, Yelp, Angi, HomeAdvisor, Thumbtack, BBB, Facebook, Apple Maps, and at least a dozen additional platforms. One inconsistency — "Suite 200" on your GBP and "#200" on Angi — creates a conflicting data signal. AI systems resolve conflicts by trusting neither listing fully and favoring a competitor whose data matches everywhere. Audit every directory, document every variation, and correct them all to match your GBP exactly.
Build a consistent review velocity system
Review count matters, but recency is the multiplier. A contractor with 40 reviews — including 8 in the last 30 days — outperforms a contractor with 120 reviews from 2022–2023 in AI recommendations. AI systems treat recent reviews as a signal that a business is actively operating and currently trusted. The highest-converting review request format: a text message sent 2–4 hours after job completion with a direct link to your Google review page. Automate this so it happens after every job without requiring manual action.
Create answer-ready content pages
Pages that directly answer the questions your customers ask give AI something concrete to extract and cite. "How much does AC replacement cost in [your city]?" "What's included in an electrical panel upgrade?" "How long does a roof replacement take?" Each of these deserves its own page that answers the question plainly in the first paragraph, then provides supporting context. Vague service pages that say "we provide high-quality plumbing services" give AI nothing to work with. Answer-ready content is especially critical for Perplexity, which surfaces and cites specific pages by URL.
Expand your citation footprint to AI-indexed aggregators
Beyond the standard directories, several aggregator platforms are directly indexed by AI systems: Houzz (home improvement), Expertise.com (local service rankings), Bark.com, and trade-specific directories like ACCA for HVAC or PHCC for plumbing. Being listed in these platforms — especially with complete profiles and positive reviews — adds AI-readable signals that competitors without those listings don't have. For local AI recommendations, breadth of credible mentions across relevant platforms is a meaningful differentiator.
Where Do ChatGPT and Perplexity Get Their Data?
Understanding the data sources is what allows you to optimize efficiently instead of guessing. ChatGPT and Perplexity each have distinct data pipelines — but both can be influenced with the same underlying actions.
| Data Source | ChatGPT | Perplexity |
|---|---|---|
| Google Business Profile | Primary signal for local recommendations | Secondary — cited less but still read |
| Website schema markup | High weight — explicit structured data | High weight — directly extractable |
| Bing index (web crawl) | Used when browsing tool is active | Primary web crawl source |
| Yelp, Angi, HomeAdvisor | Aggregated into trust score | Pages cited directly as sources |
| Review platforms | Count and recency weighted heavily | Count and recency weighted heavily |
| Answer-ready content pages | Moderate — supports structured extraction | Very high — cited by URL as source |
The practical implication: if you only have time to focus on two areas, focus on GBP completeness and schema markup. These two signals carry the most weight across both platforms and have the clearest, most direct connection to whether you get recommended. Everything else amplifies those two foundations.
What a Contractor Not Optimized for AI Looks Like
This isn't a failing business. It's a contractor with real customers, solid work, and a five-year track record — who simply hasn't made their data legible to AI systems yet.
GBP has the business category but no individual services listed. Service area field is blank.
31 Google reviews, the last one from 14 months ago. No review request system in place.
Website has no schema markup. Service pages average 150 words of generic copy.
"Street" abbreviated on GBP. "St." on Angi. "Street" spelled out on BBB. Three variations, zero matches.
No FAQ content on the site. No cost estimate pages. Nothing AI can extract and cite directly.
Not listed on Houzz, Expertise.com, or any trade-specific directory beyond the basic three.
The competitor showing up in ChatGPT results across from this contractor: 87 reviews with 14 in the last 60 days. GBP fully completed with 11 services individually named. LocalBusiness, Service, and FAQPage schema on every relevant page. NAP identical across 40+ directories. Dedicated cost and comparison pages for every major service. Three trade-specific directory profiles with complete information.
The gap isn't talent, quality of work, or advertising spend. It's data legibility. AI can't recommend what it can't confidently identify, locate, and trust. The seven-step process above is how you close that gap systematically.
Common Questions
How does ChatGPT decide which businesses to recommend?
ChatGPT builds a confidence score for each business based on the completeness and consistency of its data across multiple sources: Google Business Profile, website schema markup, review volume and recency, and citation consistency across directories. The business whose data is most complete, consistent, and recent gets the recommendation. It does not respond to paid advertising or keyword density — only trust signals.
Does Perplexity use the same signals as ChatGPT?
Both systems weight GBP completeness, schema markup, review recency, and citation consistency. Perplexity additionally places heavy emphasis on answer-ready content pages that it can cite directly by URL. If your site has pages that directly answer the questions customers ask — cost, timeline, what's included — Perplexity is more likely to surface and name you as a source.
How quickly can a contractor appear in ChatGPT results after optimizing?
GBP completeness improvements can affect AI recommendations within 2–4 weeks. Schema markup changes are typically recognized within 30 days. Citation corrections take 30–60 days to propagate across directories. Review velocity improvements compound over 60–90 days. Most contractors who complete all seven steps see meaningful improvement in AI recommendation frequency within 60–90 days.
What schema markup types matter most for contractors?
The three highest-priority types are LocalBusiness (with the correct trade sub-type: Plumber, HVACBusiness, ElectricalContractor, RoofingContractor), Service (one block per service page with service name, description, and area served), and FAQPage (on any page with a question-and-answer format). These three give AI systems the clearest possible picture of who you are, what you do, and where you serve.
How Boxi Handles ChatGPT Optimization for Contractors
Boxi is an AEO agency built specifically for local service contractors. Our Growth plan ($1,497/month) includes the complete seven-step implementation: GBP management, schema markup on every relevant page, citation repair across 40+ directories, a review velocity system integrated with your job management workflow, and answer-ready content built for your specific services and service area.
We also track AI recommendation appearances for your target queries — "best plumber in [city]," "HVAC contractor near me," "who should I call for electrical panel replacement" — so you can see your AI visibility improving month over month, not just hope it is.
If you want to know where you currently stand in ChatGPT and Perplexity results for your market, start with a free AI visibility audit. We check your GBP completion, schema coverage, citation consistency, review velocity, and current AI recommendation appearances — and give you a specific action list regardless of whether you work with us. Book a free audit →