Not all AI platforms treat small businesses equally. Our analysis of thousands of business recommendation queries across ChatGPT, Claude, Gemini, and Perplexity reveals striking differences in how each platform approaches local and small business recommendations — differences that should directly inform where you focus your AI visibility efforts.
The Research Question
We wanted to answer a specific question: among the four major AI platforms, which ones are most likely to recommend small, independent, or local businesses vs. large chains and national brands?
To investigate, we ran 10,000 local business recommendation queries across four categories — restaurants, dentists, plumbers, and software tools — in 50 US markets, ranging from large metros (New York, LA, Chicago) to mid-size cities (Austin, Nashville, Denver) to smaller markets (Springfield MO, Boise ID, Huntsville AL).
Key Findings by Platform
Perplexity: Most Small Business Friendly
Small business recommendation rate: 71%
Perplexity showed the strongest tendency to recommend independent and small businesses in local service categories. This aligns with Perplexity's architecture — it uses real-time web retrieval with explicit citations, drawing from a diverse set of sources including local directories and review platforms.
Why Perplexity favors small businesses:
- Real-time retrieval means current web presence matters more than historical brand authority
- Perplexity explicitly cites sources, and local directories (Yelp, Healthgrades) are frequently cited for local businesses
- Perplexity doesn't have strong brand preference signals in its base model — it draws from what it finds
- Content quality on a specific page can outperform general brand authority
Recommendation: Perplexity is the highest-opportunity platform for small businesses. Invest in FAQ-structured website content and Yelp presence for maximum Perplexity visibility.
Gemini: Strong for Local, but Favors Complete Profiles
Small business recommendation rate: 64%
Gemini shows strong small business recommendation rates for local queries, driven by its deep Google Business Profile integration. A well-optimized GBP is a significant equalizer against larger competitors.
Why Gemini favors complete GBP businesses:
- Gemini's local recommendation logic is heavily GBP-based
- A small dentist with 150 Google reviews and a complete GBP competes effectively with a dental chain
- Review quality and recency matter significantly
- Smaller markets show higher small business recommendation rates (less competition from chains)
Recommendation: GBP optimization is your #1 priority for Gemini visibility. Focus on review velocity, complete business profile, and location-specific schema markup.
ChatGPT: Brand Recognition Matters More
Small business recommendation rate: 48%
ChatGPT showed a stronger tendency to recommend nationally-known brands and chains compared to other platforms, particularly for software tools and some service categories. This reflects ChatGPT's training-data-based nature — well-known brands were more frequently mentioned in training data.
However, for restaurant and food categories, ChatGPT showed much higher small business recommendation rates (67%), likely because restaurant discovery content (Yelp, TripAdvisor) is so heavily represented in training data.
Why ChatGPT is more brand-aware:
- Training data over-represents major brands that receive more editorial coverage
- Without real-time retrieval, ChatGPT depends on what it learned during training
- Yelp is a major training data signal for local restaurants — equalizing small vs. large
- For non-restaurant categories, brand recognition during training is a bigger factor
Recommendation: For ChatGPT visibility, invest in Yelp (for local/restaurant) and editorial press coverage (for other categories). Building an authoritative brand over time matters more than it does on retrieval-based platforms.
Claude: Most Conservative, Highest Confidence Threshold
Small business recommendation rate: 41%
Claude had the lowest rate of small business recommendations. This isn't bias against small businesses per se — Claude is simply the most conservative AI platform about making specific recommendations. When Claude is uncertain, it often declines to recommend and instead suggests the user "check Google" or "search for local businesses."
When Claude does recommend small businesses, it tends to do so with high confidence (often citing specific credentials, reviews, or awards).
Why Claude is more conservative:
- Anthropic has prioritized accuracy over helpfulness in Claude's business recommendation behavior
- Claude often acknowledges when its information about local businesses might be outdated
- For YMYL categories (medical, legal, financial), Claude is particularly cautious
- Claude's response is often "here's how to find a good [business type]" rather than "here's a specific recommendation"
Recommendation: Claude is the hardest platform to optimize for, but the most valuable when you do appear — Claude's recommendations carry high trust authority because users know Claude is careful. Focus on authoritative directory presence and credential documentation.
Category-Level Findings
Restaurants
| Platform | Small business rate | |---|---| | Perplexity | 82% | | Gemini | 78% | | ChatGPT | 67% | | Claude | 55% |
Restaurants are the most democratic category for small businesses across all platforms. This is driven by the volume of restaurant-specific review and discovery content that all platforms have access to.
Dentists
| Platform | Small business rate | |---|---| | Perplexity | 74% | | Gemini | 68% | | ChatGPT | 45% | | Claude | 38% |
Dental shows wider variance between platforms. The Gemini premium for dentists reflects how well GBP works for dental practices.
Software Tools
| Platform | Small business rate | |---|---| | Perplexity | 52% | | Gemini | 38% | | ChatGPT | 29% | | Claude | 31% |
Software shows the lowest small business recommendation rates — understandably, as established brands like Salesforce, HubSpot, and Slack dominate training data. Smaller SaaS products must invest heavily in G2 and Capterra presence to compete.
Plumbers and Home Services
| Platform | Small business rate | |---|---| | Perplexity | 71% | | Gemini | 66% | | ChatGPT | 53% | | Claude | 43% |
Home services shows strong small business rates on all platforms, particularly in smaller markets where national chains have less presence.
Market Size Effect
A key finding: small business recommendation rates are significantly higher in smaller markets.
| Market size | Average small business recommendation rate (all platforms) | |---|---| | Large metro (>1M population) | 49% | | Mid-size city (200K-1M) | 61% | | Small city (<200K) | 74% |
In smaller markets, large chains often have weaker local review presence and less GBP optimization, while local independent businesses have built stronger community review bases. This creates a significant AI visibility opportunity for local businesses in smaller markets.
Implications for Your AI Visibility Strategy
If you're a small local business: Perplexity and Gemini are your highest-opportunity platforms. Focus on: (1) GBP optimization for Gemini, (2) FAQ-formatted website content for Perplexity, (3) Yelp presence for ChatGPT restaurant/local business queries.
If you're a small SaaS company: ChatGPT and Claude are your biggest challenges. Focus on: (1) G2 and Capterra reviews to build review-based authority, (2) Content marketing to build brand mentions in training data over time, (3) Narrow niche positioning to avoid direct comparison to established platforms.
If you're in a smaller market: Your AI visibility opportunity is proportionally larger than businesses in major metros. The competitive bar is lower, and first-mover advantage is more achievable.
Q: Does this mean small businesses shouldn't bother optimizing for ChatGPT? A: No — ChatGPT is still the largest AI platform by user count. But the optimization levers are different. For ChatGPT, focus on Yelp presence and long-term editorial citation building rather than the technical signals that work better for Perplexity and Gemini.
Q: How did you control for industries and markets in this study? A: We ran identical prompt sets across all platforms for the same categories and markets. We controlled for market size by analyzing results in five market size brackets. Some variance remains from factors like regional platform usage rates that we could not fully control for.