Research · October 2026
Twelve buying-intent prompts. Two AI assistants. One question: when a buyer asks what to buy, whose brand shows up? The answer is not a search results page. It is a paragraph. And most pet brands are not in it.
| Brand | Category | Citation share |
|---|---|---|
| Dutch | Vet telehealth | 58.3%* |
| Ollie | Fresh DTC | 39.6% |
| Sundays for Dogs | Air-dried DTC | 37.5% |
| Open Farm | Ethical | 37.5% |
| Pet Honesty | Supplements | 33.3% |
| Native Pet | Supplements | 31.3% |
| Spot & Tango | Fresh DTC | 28.6% |
| Wuffes | Supplements | 27.0% |
| Iams | Mass | 24.1% |
| Pedigree | Mass | 22.2% |
| Rachael Ray Nutrish | Mass | 21.3% |
| Beneful | Mass | 17.3% |
*Dutch measured on ChatGPT only; Gemini required sign-in.
Every brand we tested wins the prompts about itself and loses the prompts about the category. Ask “is Wuffes good” and Wuffes shows up. Ask “what is the best hip and joint supplement” and Wuffes is gone. Dasuquin and Cosequin take the answer.
Three examples.
Wuffes sells hip and joint chews. It is their flagship product. Both assistants were asked “what is the best hip and joint supplement for dogs.” Neither named Wuffes once.
Native Pet sells probiotics. Both assistants were asked “what is the best probiotic for dogs.” Native Pet got zero mentions. Purina FortiFlora owns the answer.
Sundays for Dogs sells fresh-style food. Both assistants were asked “what is the best fresh dog food delivery service.” Sundays was absent. Spot & Tango was named instead.
The health and vet prompts are worse. Across every brand we tested, the clinical questions belong to Purina Pro Plan, Hill’s Science Diet, and vet-credentialed brands like Nutramax and VetriScience. The mass brands are absent from all of them. Beneful appears only when someone asks about Beneful by name. When Gemini was asked “is Pedigree good dog food,” it answered: “Is it high quality? Generally, no.”
Citation is not luck. The brands that show up share three traits.
First, crawlers can reach them. AI crawlers are blocked by default on hundreds of top sites, often by accident. If GPTBot cannot read your site, your site cannot be cited.
Second, they publish pages that answer the question directly. AI answers quote pages that lead with the answer, back claims with sourced facts, and carry structured data. A product page is not an answer. A comparison page is.
Third, they are corroborated elsewhere. AI answers synthesize multiple sources. Brands that appear in reviews, editorial coverage, and real customer discussion get cited. Brands that exist only on their own site do not.
None of this is advertising. It is being present where the decision happens.
The next edition runs in January. Every brand on this list, plus the ones we missed, gets measured again.
If you run one of these brands, you can get your own audit before then. Twelve prompts. Your numbers against your competitors. The exact prompts where you are invisible. Free.
Method: 12 buying-intent prompts per brand, run once each on ChatGPT and Gemini in fresh chats, first answers only. Scoring: cited with a link = 2 points, mentioned = 1, absent = 0. Spot & Tango and Wuffes reflect partial ChatGPT runs (anonymous message limits cut the final prompts).