Which AI model cites small Australian businesses most often?
We ran 400 SMB-relevant prompts across GPT-4o, Claude, Gemini and Perplexity. The winner surprised us.
Most Australian SMB owners we work with have a default assumption about which AI they should care about, usually whichever one they personally use. The reality, when you look at the data, is that AI visibility is not a one-platform game. It's also not the same game across categories.
Over six weeks we ran 400 prompts across four major models, designed to mimic the kinds of questions real Australian buyers would ask before engaging a small business. The prompts spanned ten categories: tradies, accountants, financial advisers, marketing agencies, lawyers, dentists, allied health, hospitality, e-commerce, and IT services.
The methodology in one paragraph.
For each category we built 40 prompts varying in shape: "best X in suburb Y" style, problem-led ("who fixes Z"), comparison-led ("X vs Y vs Z"), and intent-led ("how do I find a good X"). We ran each prompt across GPT-4o (via ChatGPT), Claude (Anthropic), Gemini (Google), and Perplexity. Each model got the exact same prompt, same day, no context priming. We logged which Australian SMBs appeared in the answer, in what position, and with what attribution.
The headline finding.
Perplexity cited Australian SMBs most often, in 71% of relevant prompts versus a category average of 54%. ChatGPT (GPT-4o) was second at 58%. Gemini was third at 53%. Claude was last, at 34%.
The Claude number was the surprise. Anecdotally Claude is the model most people in our circle reach for. But when it comes to surfacing small specific local businesses by name, it's the most conservative; it tends to answer in general principles rather than name brands.
What it means by category.
Aggregate numbers hide the more useful detail. Some patterns by category:
- Tradies, dentists, allied health: Perplexity dominates; it pulls from Google Maps and business directories better than competitors.
- Accountants, financial advisers, lawyers: ChatGPT leads. It's surprisingly comfortable citing specific firms when they have substantive web content.
- Marketing agencies, IT services: nearly even across all four, with Claude lagging.
- Hospitality: Gemini is strongest, presumably leaning on Google reviews + Maps.
- E-commerce: ChatGPT leads, often pulling from buyer-review sites rather than the brand's own domain.
The actionable takeaway.
If you only have time to optimise for one AI platform, optimise for Perplexity. It cites more often, pulls from more sources, and is the model power-buyers increasingly use. After that, the order depends on what category you're in.
More importantly: don't optimise for the AI you use. Optimise for the AI your buyers use. We see a striking gap between the AI tooling habits of business owners (Claude / ChatGPT) and their buyers (Perplexity / Gemini / Google AI Overviews). If you only show up where you personally search, you're optimising for the wrong panel.