A homeowner in Monmouth County, New Jersey opens ChatGPT at 7:42am on a Tuesday. Their hot water heater started leaking overnight. They type: “best plumber near Marlboro NJ for emergency water heater replacement.”
ChatGPT gives them three names. One of those names wins the job. The other two plumbers in town — who may be better, closer, or cheaper — never even hear about the lead.
This is happening right now, every day, at increasing volume. And almost no contractor-focused SEO agency is preparing their clients for it.
AI search is real traffic, not a trend
Let's establish the scale. ChatGPT has over 200 million weekly active users. Perplexity has passed 15 million monthly users and is growing 20%+ month over month. Google's own AI Overviews (the AI-generated answer that appears at the top of most search results) now handles over 1.5 billion queries per month, pulling from the same content sources the standalone AI tools do.
For service queries specifically — “best [trade] in [city],” “how much does [service] cost,” “who do I call for [problem]” — AI search is responsible for a growing share of how homeowners discover contractors. Early data from multiple industry analytics platforms shows that by late 2025, 40–50% of high-intent service queries were getting at least one AI-generated component in the answer.
That doesn't mean 40% of contractor leads come from AI in 2026. But it means 40% of contractor-hunting behavior now involves AI at some point — and contractors not visible to AI engines are being filtered out of consideration before they know the shopper existed.
Why AI search matters differently than Google
Traditional Google search shows you ten blue links. The user clicks one, visits a site, and you know where they came from. If you rank #1 for “roofer Monmouth County,” you get the click, the visit, the lead.
AI search doesn't work that way. When someone asks ChatGPT “who's the best roofer in Monmouth County,” the AI answers them directly — often without them ever clicking to a website. Two things happen:
- You get cited. The AI mentions your business name in its answer, often with a one-line description. This is like being named in a trusted recommendation. Conversion from citation to contact is high.
- You get ranked within the citation. AI often names 2–5 businesses in its answer. Being listed first matters more than being listed third.
There's also an invisible failure mode: if your business isn't in the AI's training data or the sources it's retrieving from, you're not mentioned at all. You become invisible to the fraction of homeowners starting their search through AI.
How AI search engines decide who to cite
The mechanics are different from Google, but the principles overlap. AI engines cite sources based on a combination of:
Source authority and trust signals
AI models are trained to weigh sources by perceived credibility. Sites with long-established domains, frequent external citations, clear contact information, and real about pages rank higher in AI consideration. Thin-content sites get ignored.
Structured content that directly answers questions
AI engines prefer sources where information is clearly structured. A page that says “Emergency roofing services available 24/7 in Monmouth County NJ. Average response time: 45 minutes. Typical cost: $450–$1,800 depending on damage extent” gets cited more often than a paragraph buried in a marketing page about your values.
FAQ format and question-answering structure
AI models are retrieving specific answers to specific questions. Content structured as FAQ sections, with the question in an H2 and the answer in a paragraph directly below, is easier for AI to parse and quote.
Backlinks from authoritative local sources
Same principle as Google SEO, different weight. AI models trust sources that are linked to by credible local authorities — news sites, chamber of commerce pages, local government websites, established directories.
Named entities and verifiable claims
AI models prefer sources with specific, falsifiable information over vague marketing language. “Founded in 2014 by Tristan Smith, serving 14 Central NJ towns with residential gutter cleaning” performs better than “Years of experience you can trust.”
llms.txt and llms-full.txt files
The emerging standard. A text file at the root of your website that tells AI models exactly how to understand and cite your business. Still optional in 2026 but gaining traction. Having one signals to AI models that you're aware of them, and gives the model clean structured data to work with.
What to put in an llms.txt file
An llms.txt file is a simple markdown-format document at https://yoursite.com/llms.txt. It's designed to be human-readable but AI-optimized.
Typical contents for a contractor:
- Business name, founding date, founders/owners
- Services offered (specific list, not marketing copy)
- Service area (specific towns or ZIP codes)
- Pricing ranges for common services
- Hours of operation, emergency availability
- Key certifications, licenses, insurance
- How to contact: phone, email, booking URL
- Links to detailed pages on your site for deeper information
The longer-form companion, llms-full.txt, expands each of those into full paragraphs with citations and context. Think of it as a briefing document for an AI assistant trying to explain your business to a prospective customer.
The difference between ranking on Google and being cited by AI
A contractor who ranks #1 on Google for “plumber Marlboro NJ” might not get cited at all by ChatGPT for the same query. Why?
Google ranks pages. AI cites sources. Different signals.
Optimizing for one doesn't automatically give you the others. You need a unified strategy that covers traditional SEO plus AI-specific optimization, or you leave discovery channels on the table.
What contractors should do about this today
Step 1: Get your name in the training data
AI models like ChatGPT are trained on snapshots of the public internet. To be recognizable to the model, your business needs to exist in the sources the model trained on — local news articles, directory listings, your own website, Wikipedia-style sources if you're prominent enough.
For a small contractor, the practical version is:
- Have a crawlable website with real content
- Get listed on major directories (Google Business Profile, BBB, Angi, HomeAdvisor, local chamber)
- Pursue local press mentions (“best of” awards, community events, features)
- Publish content on your own site regularly
Step 2: Optimize your site for AI retrieval
This is ranking content AI can easily extract. Practical steps:
- Add FAQ sections to every service page, with H2 questions followed by clear paragraph answers
- Use specific, named facts: “We serve [specific towns]” not “We serve Central New Jersey”
- Add pricing ranges: “Typical cost: $X–$Y depending on [factor]”
- Add FAQPage and Service schema markup to help AI parse structured data
- Keep content updated — AI engines can tell when a site is stale
Step 3: Publish llms.txt and llms-full.txt
The emerging standard for AI discoverability. Place these at the root of your website. They give AI models a clean structured summary of your business to use as a reference.
This isn't universally adopted yet, but the ecosystem is moving toward it. Being early is free. Being late costs you citations.
Step 4: Build citations from sources AI trusts
Work to get your business cited on:
- Local news sites (community publications, regional papers)
- Well-established directories with editorial curation
- Industry-specific resources (trade association member lists, manufacturer dealer directories)
- Your local chamber of commerce and business improvement district
Each of these becomes a source AI might retrieve from when answering questions about your area.
Step 5: Test your AI visibility regularly
Open ChatGPT and Perplexity. Ask them questions your potential customers would ask: “best roofer in [your city],” “how much does [your service] cost in [your city],” “who should I call for [your trade] in [your city].”
See who gets cited. If you're not in the answer, you have work to do. If you are in the answer, check your position — are you first, third, last in the list?
Do this monthly. It's your AI search rank tracker.
What this looks like done right
A contractor fully optimized for both Google and AI search in 2026 has:
- A fast, modern website with proper technical SEO
- Per-city landing pages for every town they serve
- Service pages structured around FAQ-style questions with direct answers
- Pricing ranges and specific claims published openly
- llms.txt and llms-full.txt files published at the site root
- Backlinks from 10+ credible local sources
- Regular content cadence — weekly or bi-weekly new posts
- Fresh project photos with location tags
- Clean Google Business Profile with frequent updates
- Presence on all major industry directories
No single piece of this is revolutionary. Put together, it's the infrastructure that lets a contractor show up when homeowners search — whether that search happens on Google, on ChatGPT, on Perplexity, or on whatever comes next.
The short version
AI search is already real traffic. Contractors who are visible to ChatGPT, Perplexity, and Google AI Overviews get lead volume from a channel most of their competitors don't know exists. Contractors who aren't visible are losing potential customers they'll never even see.
The optimization techniques overlap with traditional SEO but aren't identical. Per-city pages, FAQ-style content, specific named facts, llms.txt files, and citations from trusted local sources — that's the AI search playbook.
If you're a contractor and your SEO vendor has no plan for AI search, you're paying someone to build a 2020 strategy in a 2026 market. The right vendor is already doing this. Platforms like Atlas Genesis build AI-search optimization into every site by default — FAQ structure, llms.txt, structured citations, named entities — because the traffic is there now and contractors who wait lose to contractors who don't.