What AI assistants say about Golden Retriever breeders in Arizona
On 2 September 2026 I asked Claude, ChatGPT and Perplexity the same question — who are the best Golden Retriever breeders in AZ? — and kept every word of all three answers. Between them they named twenty-two Arizona programmes. Only one was named by all three. This page reports what they said, who they read to decide, and nothing else.
What this page is, and what it is not
This is not a recommendation list. I have not visited these kennels, met their dogs, or verified a single health clearance. Nothing here is an endorsement by me or by Direct-Me, and a programme appearing on this page is not evidence that it is good — only evidence that a machine said its name on a particular day.
What it is: a dated, quoted record of what three AI assistants tell an Arizona family who asks them where to buy a Golden Retriever. That is worth writing down because it is what a growing share of buyers now read before they ever reach a breeder’s website, and because almost nobody has checked what it actually says.
The headline finding: they barely agree
Twenty-two distinct Arizona programmes were named across the three answers.
| Named by | Programmes |
|---|---|
| All three assistants | 1 |
| Exactly two | 9 |
| Only one | 12 |
More than half the programmes named were named by a single assistant. If you ask one AI which Arizona breeders to consider and act on the answer, you are acting on a list that the other two largely do not share.
The only programme named by all three was Heavenly Sunsets Golden Retrievers. I make no claim about why, and the answers themselves do not agree on a reason.
Named by two of the three
Chiricahua Retrievers · Cosmic Gold Retrievers · Goldminer Goldens · Oro Valley Golden Retrievers · Suzy’s Golden Girls · Valor Goldens · Westwing Retrievers · Wildfire Golden Retrievers · Xanadu Goldens
The three answers were different in kind, not just in content
Claude declined to rank at all. Its answer opens by saying that “best” depends on what the buyer wants — show lines, field lines, service temperament, English cream versus American — and then points first at the two GRCA-affiliated Arizona clubs before naming any kennel. It named seven, and spent most of its answer on how to vet a breeder rather than which one to pick.
ChatGPT ranked, with star ratings, and named fifteen. The ratings were not its own: they came from a third-party roundup site, which it cited repeatedly.
Perplexity listed eleven, in a table, with a location note for each and no ratings.
Where each assistant got it — and this is the part worth acting on
The three do not share a gatekeeper. Each leaned on a different source, and for one of them that dependence was almost total.
| Assistant | What it leaned on |
|---|---|
| ChatGPT | 16 of its 21 sources were a single domain, devotedtodog.com. The star ratings in its answer are that site’s ratings. |
| Perplexity | Leaned hardest on gooddog.com (8 citations), then individual breeder websites and a breed-club directory. |
| Claude | The most spread out: breeders’ own sites, a Golden Retriever forum, the Valle del Sol club site, and devotedtodog.com — none dominant. |
If you breed Goldens in Arizona, that table is the actionable part of this page. Whether an assistant names you is largely downstream of whether you appear on the handful of sources it happens to read, and those sources are different for each one. Getting onto the list that ChatGPT reads does very little for what Perplexity says.
How this was run
- Date: 2 September 2026. Assistants change their answers; this is a snapshot, not a standing fact.
- Question, asked verbatim to each: “Who are the best Golden Retriever breeders in AZ?”
- Assistants: Claude (Opus), ChatGPT (GPT-4.1) and Perplexity (Sonar), each with web search enabled, each asked once, in one sitting.
- Sources counted: 123 citations across the three answers.
- Names: read out of the answer text by a person. No name on this page was produced by a matcher or inferred from a domain.
This page is one cell of a larger run — the same five questions across twenty-five states and sixteen breeds, all of it recorded. The method, and the figures that came out of the first twelve states, are in the Breed Study. The raw data is published as CSV.
If you are named here and something is wrong
These are other people’s businesses and the assistants are not always right about them. If a detail attributed to you is inaccurate, or you would rather not appear on this page at all, write to me and I will correct or remove it. I will also tell you plainly what the assistants said, in full, so you can see the source of the error rather than take my word for it.
I breed Goldens and Cavaliers myself, in Oregon. I ran the same questions against my own programmes before I ran them against anyone else’s, and the answer was not flattering — three mentions, all in my own state. That is why I started counting.