Real Examples of ChatGPT Recommending Local Businesses

There's a lot of theory about AI visibility. But what does it actually look like when ChatGPT recommends a business? What exactly does it say? And what patterns emerge across different industries and query types?

I've been documenting real ChatGPT sessions, testing queries across industries and locations. This article shares actual examples—not hypotheticals—of what AI recommendations look like in practice.

Example 1: "Best Italian restaurant in Soho, London"

Why these businesses?

I analysed the online presence of both recommended restaurants:

• Cross-platform presence — Both have reviews on Google, TripAdvisor, Time Out, and multiple food blogs
• Media coverage — Featured in The Guardian, Evening Standard, and Eater London
• Specific positioning — "Regional Italian" and "Italian-American with fresh pasta" are clear differentiators
• Quotable qualities — "Watch the chefs at work", "pasta made in-house"—these are specific, memorable details

What ChatGPT didn't recommend

Notably absent: generic Italian restaurants with fewer reviews, chains, and restaurants without clear differentiation. One nearby Italian restaurant with similar Google ratings wasn't mentioned—their website positioned them simply as "Italian cuisine" without further specificity.

Example 2: "Good accountant for freelancers in Bristol"

Analysis

• Niche specialisation wins — Gorilla Accounting explicitly positions itself for contractors/freelancers. This directly matches the query
• Clear pricing model — "Fixed-fee pricing" is mentioned because it's stated clearly on their website and in reviews
• Online platform presence — Crunch is mentioned partly because their software-plus-accountancy model is distinctive and well-documented
• Third option hedging — ChatGPT often includes a "if you prefer X" alternative, showing it considers different user preferences

Example 3: "Best tattoo artist for fine line work in Manchester"

What made the difference

• Style-specific positioning — These studios explicitly mention "fine line" on their websites and in their Instagram bios
• Portfolio visibility — Their work is documented and findable on Instagram, which AI can reference
• Helpful caveat — Notice ChatGPT adds value by advising to check individual artist portfolios. It's not just answering—it's being helpful

Example 4: "Recommend a wedding photographer in Edinburgh"

Pattern recognition

Each recommendation came with a style descriptor:

• "Documentary-style"
• "Fine art approach"
• "Classic style with high-end production"

These descriptors came directly from the photographers' websites and reviews. They had clearly articulated their style in ways that AI could understand and repeat.

Example 5: "Good co-working space with private offices in East London"

Interesting observations

• Mix of sizes — ChatGPT includes both established players (WeWork) and niche options (PLATF9RM)
• Specific features quoted — "Natural light and design", "designed for growing tech companies"—these are from their marketing materials
• User preference awareness — "If you want something smaller and more community-focused"—ChatGPT anticipates different needs

Example 6: "Specialist running shoe store in London"

Why not a generic sports store?

The query specified "specialist"—and ChatGPT understood that. Generic sports retailers like JD Sports or Sports Direct weren't mentioned, despite having more locations and higher brand recognition. The AI matched the specificity of the query to businesses with matching positioning.

Patterns Across All Examples

Looking across these examples, several patterns emerge:

1. Specificity matches

Businesses get recommended when their positioning closely matches the query. "Fine line tattoo artist" → studio that explicitly mentions fine line work. "Accountant for freelancers" → accountant that explicitly serves freelancers.

2. Quotable differentiation

Every recommendation came with a reason. These reasons were pulled from:

• Business websites (positioning statements, feature descriptions)
• Reviews (commonly mentioned qualities)
• Press coverage (distinctive features highlighted in media)

3. Multi-source validation

Recommended businesses appeared across multiple platforms: their own website, review sites, media coverage, social media. This cross-validation gives AI confidence.

4. Three-recommendation pattern

ChatGPT typically recommends 2-4 businesses, often with different options for different preferences. This suggests businesses are competing for limited slots—if you're not in the top 3-4 for your query, you're invisible.

5. Helpful context

AI adds value beyond just naming businesses—it explains why, suggests what to look for, and anticipates different needs. Businesses that provide content making this easy (FAQs, comparisons, guides) are more likely to be featured.

What Doesn't Appear in AI Recommendations

Equally instructive is what ChatGPT doesn't recommend:

• Businesses with generic positioning — "Full-service agency", "serving all clients", "your one-stop shop"
• Businesses with thin online presence — Website only, no reviews, no third-party mentions
• Businesses optimised only for paid ads — Google Ads don't influence organic AI recommendations
• Businesses with outdated information — Closed locations, old contact details, stale content
• Businesses blocking AI crawlers — Some websites inadvertently block GPTBot in robots.txt

How to Get Into the Recommendation Set

Based on these real examples, here's what you need:

1. Clear, specific positioning
   Match the queries your potential customers actually ask. If someone asks for "X in Y", you need to explicitly be positioned as X in Y.
2. Quotable differentiation
   What can AI say about you? Provide distinctive features, specialisations, or approaches that can be stated concisely.
3. Multi-platform presence
   Be findable on your website, Google Business Profile, industry directories, review platforms, and ideally in media/publications.
4. Consistent, current information
   Same business name, address, and positioning everywhere. Recent reviews and updated content.
5. Earn reviews that mention your positioning
   Reviews that say "great for [specific thing]" reinforce your positioning signal.

This Article Will Be Updated

AI behavior changes as models update. I'll continue testing and documenting real examples. If you've noticed interesting patterns in how AI recommends (or doesn't recommend) businesses in your industry, I'd love to hear about it.

The goal is to build a body of evidence—real, documented examples—so business owners can understand what AI visibility actually looks like in practice, not just in theory.