We've written before about the four UK golf operators we tested against the same buyer prompts on Perplexity and ChatGPT: St Mellion Estate, Redlibbets Golf Club, Orpington Golf Centre and Nizels Golf & Country Club, findings first covered by The Golf Business. Looking back across that result set, one pattern stands out clearly enough to deserve its own closer look: Perplexity was consistently the weakest platform for every single operator tested.
| Club | Perplexity result | ChatGPT result |
|---|---|---|
| St Mellion Estate | Not mentioned | Ranked 4th of 4 |
| Redlibbets Golf Club | Not mentioned | Ranked 5th of 6 |
| Orpington Golf Centre | Afterthought mention | Not mentioned |
| Nizels Golf & Country Club | Ranked last | Ranked last |
Two of the four were absent from Perplexity entirely. A third only appeared as an afterthought. The fourth was ranked last. Not one of the four came back stronger on Perplexity than on ChatGPT. That's not four unrelated results, it's a pattern worth understanding on its own terms.
Why Perplexity behaves differently
Perplexity positions itself as an answer engine built on live web search and citation, rather than a model reasoning primarily from training data. In principle that should favour businesses with strong current web presence. In practice, our testing suggests it leans more heavily on a narrower set of sources it treats as authoritative for local recommendations, and it appears more literal about matching the specific wording of a query to specific wording it can find in its indexed sources, compared to ChatGPT or Gemini's broader interpretation of intent.
The practical effect is that a business can have strong general web authority and still be functionally invisible to Perplexity if the specific sources it's drawing from for that query type don't happen to include or favour that business. This isn't a quality signal. It's a sourcing behaviour, and it's one that plain SEO strength doesn't automatically fix.
What this means commercially
If your business only checks one AI platform and assumes the result generalises, Perplexity is the one most likely to give you a false negative or a false positive depending on which platform you happened to test. A business that looks strong on ChatGPT could be functionally absent on Perplexity for the same customer question, with no way to know unless it's tested directly.
What tends to help
Based on the pattern across these results and the wider set of sites we test, a few things consistently correlate with stronger Perplexity performance: a Google Business Profile with a healthy volume of recent, specific reviews (Perplexity appears to weight review platforms heavily as a citation source), consistent business details across every listing and directory, and content on your own site specific enough to be directly quotable rather than generic marketing copy. None of these are Perplexity specific fixes exactly, they're the fundamentals done properly, but Perplexity seems to reward them, or punish their absence, more sharply than the other platforms do.
The takeaway
A single weak result on one platform can look like noise. Four for four is a pattern. If you've only ever checked ChatGPT, or never checked any AI platform at all, Perplexity is the one most likely to be quietly telling a different story about your business than you assume.