💡 Key Insight: Query match OUTRANKS star rating in AI answers. A 3.3-star car wash won because it matched "no-touch SUV clearance 24/7." Reviews on Google/Yelp listings do NOT feed AI answers unless republished to crawlable surfaces.
📚 Lesson: How AI Now Finds Local Businesses
The Behavior Shift
GatherUp consumer data (fall 2025):
- 55% of consumers consulted Google/Bing AI summaries
- 48% asked ChatGPT about a local business
- 31% asked multiple times
Query Match > Star Rating
When Annie Jackson asked Google for "no-touch car wash that fits an SUV in Norfolk, VA," Google returned a 3.3-star business. The answer included clearance height and 24/7 hours above the star rating.
"Query match outranked rating."
Reviews Don't Feed AI Answers (Unless You Republish)
Critical finding from Jason Wertham:
- Google, Yelp, and major directories block LLM crawlers from reading review content on listings
- Reviews on listings support rankings and conversion there, but do NOT enter AI answers directly
- When you republish reviews to public social channels or embed them in widgets on your own site, they become "fair game for LLM tools"
Review Recency > Star Rating
- 45% of users prioritize review recency over star rating
- 60% trust detailed written reviews over rating-only
- 70% prefer review request within 72 hours of transaction
- Consumers switch from "most relevant" to "newest" sort because recent reviews predict their experience
🛠️ Practice Activity: Run the 4-Prompt Audit
Prompt 1: Brand Name Query
Ask ChatGPT/Perplexity:
"What do you know about [Your Brand Name]? List 3 things you believe are true."
What to check: Is the information accurate? Is the tone positive, neutral, or negative? Does it mention services/locations correctly?
Prompt 2: Service + Location Query
Ask:
"I'm looking for [service] in [city]. What are the best options?"
What to check: Does your brand appear? What does the AI say about you? What sources does it cite?
Prompt 3: Comparison Query
Ask:
"Compare [Your Brand] vs [Competitor] for [service] in [city]."
What to check: How does the AI describe each? Does it mention your differentiators (e.g., "local agent," "bilingual," "free consultation")?
Prompt 4: Location Spot Check
Ask:
"Tell me about [Your Brand] at [specific address]. What do people say?"
What to check: Is address/hours correct? What review sentiment does the AI convey? Does it mention any issues?
Document Your Findings
- Screenshot each AI response
- Note inaccuracies, gaps, or negative framing
- Identify which sources the AI cites (your site, social, directories, third-party mentions)
- Compare AI description to your desired brand positioning
🔧 Implementation Task
Site: xactinsure.net + social channels
Task: Deploy crawlable review widget + republish top reviews
Step 1: Choose Crawlable Review Display
Options for making reviews crawlable by LLMs:
- Review widget on your own site: Elfsight, Trustpilot widget, or custom embed, ensure it's in plain HTML (not iframe-only)
- Social republishing: Post client testimonials to Facebook, Instagram, LinkedIn with business tags
- Blog/Case study format: "Client Story: How [Name] Found the Right Medicare Plan" with full review text
Step 2: 72-Hour Response Window
- Set GHL automation to request review within 24 hours of enrollment confirmation
- Set GHL automation to alert team if review goes unanswered for 48 hours
- Assign Carmela or Bella to respond to all new reviews within 72 hours
- Include business reply with review when republishing (shows active management)
Deadline: August 4, 2026 (1 week)
Assign to: Catherine (widget + site) + Carmela (response workflow) + Jen (social republishing)
✅ WITNESS Check
Before marking complete, verify:
- 4-prompt audit completed for XactInsure (screenshots saved)
- At least 10 top reviews republished to crawlable surface (site widget or social)
- Review widget on xactinsure.net renders in plain HTML (inspect source, confirm text is in DOM)
- 72-hour response workflow active (test with a sample review)
- Team knows how to respond to negative reviews per SOP