ChatGPT, Gemini, and Perplexity now recommend local businesses to users asking 'best plumber in Houston' or 'what's a good dentist in Tampa'. Here's how to optimize for AI assistant recommendations , the new frontier of local visibility.
In this guide
How AI assistants choose recommendations
- Crawl public web data. Your website, GBP, citations, blog content all feed into the model.
- Weigh authority signals. Wikipedia, BBB, official directories increase trust score.
- Extract from structured data. Schema markup is heavily weighted , FAQ, LocalBusiness, Service.
- Cross-reference entity consistency. Same name, address, services across all web sources.
- Pull from FAQ content. Direct question-answer formats are AI extraction gold.
How to optimize for AI recommendations
Step 1: Deploy llms.txt at your domain root with canonical business facts. Step 2: Add comprehensive FAQ schema to every key page. Step 3: Ensure entity consistency across web (NAP, descriptions). Step 4: Use specific data in content (prices, percentages, timelines). Step 5: Build authority through legitimate citations and content.
Patterns that fail with AI
Vague marketing language ('we're the best'). Inconsistent business names across platforms. Missing schema markup. Thin content with no specific data. Outdated information across sources.