How to find out where you stand right now
Go to ChatGPT right now. Type:
"Who's the best [your trade] in [your city]?"
Look at the answer. Is your business mentioned? If not, you have a GEO problem — and it's costing you.
Repeat the test in Perplexity, Google's AI Overview, and Claude. Track which businesses get cited in each.
In our experience auditing client AI visibility across Texas markets, the patterns are clear: the businesses cited by AI are not always the businesses ranking highest in traditional Google search. AI uses different signals. Optimizing for AI search is its own discipline.
How AI models decide which businesses to recommend
When a user asks an AI for "the best [trade] in [city]," the model doesn't have a database of trade businesses ranked by quality. It generates an answer based on its training data and (for newer models) live web search results.
Six factors push a business into the AI's answer:
1. Review density and sentiment across multiple platforms
AI models cross-reference Google reviews, Yelp, BBB, industry directories, and forum mentions. A business with 200 Google reviews + 50 Yelp + 30 BBB + mentions on Nextdoor and Reddit is "knowable" in a way a business with reviews on only one platform isn't.
The sentiment in those reviews matters too. AI parses the actual text — businesses described as "reliable," "honest," "fairly priced" rank above businesses with positive star ratings but generic content.
2. Website content depth and authority
Thin sites (homepage, contact, maybe a services page) almost never get cited. AI models prefer sites that have demonstrably deep coverage of their topic.
For a roofing business: dedicated pages for each service, comprehensive FAQ content, pricing guides, location pages, blog content covering insurance claims, storm response, material comparisons, maintenance, financing. The more your site demonstrates expertise, the more likely it is to be cited.
3. Structured data / schema markup
This one's underappreciated. Schema markup is JSON-LD code embedded in your pages that tells parsers (Google, AI models, social media previews) exactly what your business is and what's on the page.
For trade businesses, the schemas that matter most:
LocalBusiness(or more specific:Plumber,RoofingContractor,HVACBusiness)Service(one per major service you offer)FAQPage(on every service page)ReviewandAggregateRating(review stars and counts)HowTo(for instructional content)BreadcrumbList(navigation context)
AI models parse this directly. A site with comprehensive schema is dramatically more "extractable" than one without.
4. Directory presence and NAP consistency
Your business needs to appear, with consistent Name/Address/Phone, on the major directories AI training data references. That includes:
- Google Business Profile
- Yelp
- BBB
- Apple Business Connect
- Bing Places
- Industry-specific (HomeAdvisor, Angi, Houzz, BuildZoom)
- Local chamber of commerce
Inconsistent NAP across these listings reduces AI's confidence in your business. The same business listed as "Mcanix LLC," "MCANIX," and "Mcanix Marketing" looks like three different (less trustworthy) entities to a parser.
5. Brand mentions across the web
This is the hardest part to influence but the most powerful: mentions of your business in news articles, trade publications, forum discussions (Reddit, Nextdoor), social media, blog posts on industry sites.
AI models build a knowledge graph from how often and where your brand is mentioned. A roofer mentioned in three local news stories about storm response is more "authoritative" than one mentioned nowhere.
Tactical ways to build brand mentions:
- Local PR: pitch local journalists on stories where you have unique insight (storm season prep, energy efficiency, etc.)
- Industry publications: contribute to trade pubs like Roofing Contractor, ACHR News, Contractor Magazine
- Sponsorships: local sports teams, charity events — these get press coverage that gets indexed
- Founder content: an active LinkedIn presence from the business owner adds branded mentions to AI training data over time
6. Direct answers to common questions
AI models pull complete, quotable sentences. Content that directly answers a question in clear, factual language outranks content that buries the answer.
Compare:
"We've been serving the Dallas community for over a decade, and our team is committed to providing exceptional roofing services with the highest quality materials and craftsmanship."
vs.
"A typical asphalt shingle roof replacement on a 2,000 sqft home in Dallas costs $8,500-$14,000 in 2026 and takes 1-2 days to complete."
The second one gets cited. The first one is filler.
The step-by-step playbook
Here's the order of operations for a trade business serious about AI visibility:
Step 1: Audit your current AI visibility
Test in ChatGPT, Perplexity, Claude, and Google AI Overview. Track which competitors are cited where. This baseline gives you something to measure against.
Step 2: Fix the foundation
- Claim every relevant directory listing
- Standardize NAP across all platforms (we recommend an audit via BrightLocal or Whitespark)
- Get to 100+ Google reviews with a steady velocity of new reviews
Step 3: Build authoritative content
- Comprehensive service pages (one per service)
- Detailed FAQ pages (6-10 questions per service)
- Pricing guides for your market
- How-to content (when appropriate to your trade)
Step 4: Add schema markup everywhere
- LocalBusiness on every page
- Service on each service page
- FAQPage on every FAQ section
- BreadcrumbList for navigation context
- AggregateRating with current Google rating + review count
Use Google's Rich Results Test to validate.
Step 5: Get mentioned beyond your own website
- Local press, trade publications, guest posts
- Active social/LinkedIn presence from leadership
- Industry forum participation (carefully — not spam)
- Sponsorships with local events that get coverage
Step 6: Monitor monthly
- Quarterly audit of your AI visibility for top queries
- Adjust strategy based on what's working
- Track competitor movement
Why this matters NOW
In our experience, the trade business marketing space is 12-18 months behind in adopting GEO. Most agencies don't offer it yet. Most business owners haven't heard of it.
That's the opportunity. The businesses that move now will own AI search results for their market by the time their competitors realize the game changed. Once everyone catches up — and they will — the playing field levels.
This is similar to the early days of local SEO (2010-2014). The plumbers and roofers who optimized their GBP and built citations early ended up dominant. The ones who waited spent the next decade trying to catch up.
What doesn't work
- Paying for "AI listings." No service legitimately puts you in ChatGPT's training data. Scam.
- Stuffing your site with AI-generated content. Detection is good, and the content itself is usually too thin to be cited.
- Treating GEO as a one-time project. Like SEO, AI visibility is ongoing work. Training data updates. Algorithms change. Competition shifts.
The bottom line
AI search is replacing traditional search faster than most trade businesses realize. The companies cited by ChatGPT, Perplexity, and Google AI Overview have an enormous advantage — they show up in answers their competitors don't.
The good news: the work isn't mysterious. It's a stack of fundamentals — reviews, schema, content depth, directory consistency, brand mentions — done with intentional structure for AI extraction.
If you want a free AI visibility audit (we'll check where you show up in ChatGPT, Perplexity, and Google AI Overview for your top 20 queries, and tell you exactly where to start), get in touch. For more context, see our intro to GEO and the impact of Google AI Overviews.