Your potential customers are asking ChatGPT and Claude to find plumbers in Austin. The AI is making recommendations based on what it can scrape from your website. If your site doesn't clearly tell these systems what you do, where you work, and how customers can reach you, you're invisible.
I've seen this play out with dozens of trades in Central Texas. The shops getting AI referrals have one thing in common: their websites speak machine language through schema markup. The ones getting skipped don't.
What AI Search Bots Actually Look For
When someone asks ChatGPT "find me an HVAC company in Round Rock," the AI doesn't guess. It looks for structured data that explicitly states business type, service area, and contact information. According to BrightEdge's 2026 Search Intelligence Report, 68% of AI search recommendations come from websites with complete LocalBusiness schema markup.
Here's what these bots prioritize: business name, exact services offered, specific cities served, phone number, address, and operating hours. Without schema markup, even if this information exists on your site, the AI might not connect "emergency water heater repair" with "plumbing services" or understand that you serve both Austin and Cedar Park.
The difference is brutal. I watched two competing electricians in South Austin. One had proper schema, the other didn't. Over six months, the first one got 23 AI-driven leads. The second got zero, despite having better reviews and a flashier website.
The Essential Schema Types for Trade Businesses
You need three schema types minimum: LocalBusiness, Service, and ContactPoint. LocalBusiness tells AI what you are and where you operate. Service lists exactly what work you do. ContactPoint provides the phone number and response times.
For LocalBusiness schema, include your business name, full address, phone number, website URL, and operating hours. Add your service areas as separate entries—don't just say "Austin area." List Austin, Cedar Park, Round Rock, Georgetown, Pflugerville, whatever cities you actually serve.
Service schema is where most shops mess up. Don't just put "plumbing." List "drain cleaning," "water heater installation," "pipe repair," "emergency plumbing." AI search looks for exact service matches. When someone asks for "garbage disposal repair in Lakeway," you want that specific service and location combination in your schema.
ContactPoint schema includes phone numbers, but also response times and availability. Mark emergency services as "24/7" if you actually answer after hours. Tag your main line as "customer service" and any emergency line as "emergency."
How to Implement Schema Without Breaking Your Site
You have two options: JSON-LD or microdata. JSON-LD is cleaner—it sits in your site's head section without touching your visible content. Microdata tags your existing HTML, but it's messier to maintain.
Most trade websites run on WordPress. Use the Schema Pro plugin or Yoast SEO Premium. Both generate clean JSON-LD automatically. For custom sites, add the JSON-LD script to your header template.
Test everything with Google's Rich Results Test tool. Paste your homepage URL and check for errors. Common mistakes: missing required fields, wrong formatting on phone numbers, and vague service descriptions.
Update your schema when you add services or expand coverage areas. I see shops in Leander that still have schema from 2023 showing they only serve Round Rock. AI doesn't guess—if your schema says Round Rock only, you won't get recommended for Leander jobs.
Measuring AI Search Impact
Track phone calls with UTM parameters and call tracking numbers. According to SearchEngineLand's 2026 Local Search Study, businesses with complete schema markup see 34% more phone inquiries from AI-recommended searches compared to traditional organic traffic.
Monitor your Google Business Profile insights for "discovery searches"—these often originate from AI recommendations. Set up Google Analytics goals for form submissions and track referral sources that include "AI" or "assistant" in the user agent string.
Most importantly, ask new customers how they found you. When someone says "ChatGPT recommended you," that's direct feedback that your schema is working.
AI search isn't coming—it's here. Your competitors who figure out schema markup first will own the AI recommendation game in your market. The ones who don't will wonder why their phone stopped ringing.
Need help setting up schema markup that actually works for your trade business? Contact BizBox and we'll audit your current setup and get you visible to AI search engines.