A 1,200-response audit of how Claude, ChatGPT, Gemini and Perplexity construct the category—and which brands enter the recommendation set.
The audit asked unbranded questions that a prospective buyer might use when choosing an extra virgin olive oil. It did not ask the platforms what they knew about any named brand.
Fifty prompts were organized across five commercial-intent clusters. Each prompt was run three times in its original form and three times with controlled query variation across four AI platforms, producing 1,200 response trials. All 1,200 completed successfully.
California Olive Ranch appeared in 645 of 1,200 responses, nearly twice as often as second-place Cobram Estate. It is the only brand that leads every query cluster and remains strong across all four platforms.
California Olive Ranch reaches health, flavor, cooking, origin and purchase intent on every platform. Most competitors achieve relevance through a narrower combination: a particular platform, a specific use case or a stronger digital purchase layer.
A brand can look established in the combined results while depending heavily on one platform. Counts below show how many of each platform’s 300 responses included the brand.
| # | Brand | Claude | ChatGPT | Gemini | Perplexity | Total | Inclusion |
|---|---|---|---|---|---|---|---|
| 1 | California Olive Ranch | 226 | 208 | 115 | 96 | 645 | 53.8% |
| 2 | Cobram Estate | 168 | 49 | 55 | 56 | 328 | 27.3% |
| 3 | Kirkland Signature | 114 | 112 | 18 | 30 | 274 | 22.8% |
| 4 | Brightland | 121 | 24 | 72 | 33 | 250 | 20.8% |
| 5 | Colavita | 4 | 146 | 41 | 54 | 245 | 20.4% |
| 6 | Graza | 67 | 10 | 106 | 40 | 223 | 18.6% |
| 7 | Lucini | 61 | 86 | 51 | 24 | 222 | 18.5% |
| 8 | Kosterina | 99 | 19 | 66 | 19 | 203 | 16.9% |
| 9 | Castillo de Canena | 78 | 76 | 31 | 15 | 200 | 16.7% |
| 10 | Laudemio Frescobaldi | 76 | 64 | 30 | 5 | 175 | 14.6% |
| 11 | Bono | 0 | 4 | 129 | 38 | 171 | 14.2% |
| 12 | Partanna | 18 | 34 | 93 | 24 | 169 | 14.1% |
| 13 | Olivea | 0 | 0 | 93 | 70 | 163 | 13.6% |
| 14 | Olio Verde | 68 | 59 | 1 | 9 | 137 | 11.4% |
| 15 | La Tourangelle | 1 | 57 | 61 | 9 | 128 | 10.7% |
Red indicates zero appearances; amber indicates five or fewer. The total is out of 1,200 response trials. The platform columns are each out of 300.
The open-category correction changes the published top 15: Bono enters at #11 and Olio Verde at #14. Both surfaced organically in the raw answers but were excluded from the legacy 50-brand CSV. Brightland remains #4.
Gemini named the most normalized brands per response; Perplexity named the fewest. Gemini used web grounding in all 300 trials recorded in the audit metadata. ChatGPT used web retrieval in 24 of 300, concentrated most heavily in Retail, DTC & Gifting. Retrieval status was not available in comparable metadata for Claude or Perplexity, so no cross-platform retrieval-rate claim is made for those systems.
These brands illustrate the strongest platform dependencies in or near the leading group. “Largest-platform share” is the portion of a brand’s total appearances supplied by a single AI platform.
| Brand | Strongest platform | Weakest platform | Largest-platform share | Total appearances |
|---|---|---|---|---|
| Olivea | Gemini (93) | Claude (0) | 57.1% | 163 |
| Colavita | ChatGPT (146) | Claude (4) | 59.6% | 245 |
| Carapelli | Perplexity (74) | ChatGPT (0) | 74.7% | 99 |
| La Tourangelle | Gemini (61) | Claude (1) | 47.7% | 128 |
| Rincón de la Subbética | Perplexity (45) | Claude (1) | 60.8% | 74 |
| Iliada | Claude (56) | ChatGPT (1) | 91.8% | 61 |
Olivea is the clearest retrieval-led visibility case: 163 appearances, all from Gemini and Perplexity. By contrast, Iliada’s 61 appearances come overwhelmingly from Claude. Neither pattern represents category-wide authority.
Each cluster contains 240 response trials: 10 prompts × three runs × two query forms × four platforms.
Health & Functional
Health claims, polyphenols, freshness and everyday wellness
California Olive Ranch leads this cluster across all four platforms. Olivea ties Cobram Estate for third on total appearances, but its visibility comes entirely from Gemini and Perplexity—evidence that strong retrievable health content can create rapid visibility without broad model familiarity.
Flavor & Sensory Quality
Pepperiness, fruitiness, finishing quality and sensory character
The leader’s margin remains decisive, but the next four brands are separated by only three appearances. Familiarity and broad distribution help Cobram Estate, Colavita and Kirkland Signature, while Partanna’s distinctive Sicilian and peppery associations give it a more specific sensory foothold.
Cooking & Culinary Use
High-heat cooking, finishing, dressings and practical versatility
This is California Olive Ranch’s strongest cluster. Kirkland Signature follows on everyday utility and value, while Colavita and Lucini tie for third. Graza reaches the top five despite receiving no ChatGPT appearances in this cluster, showing that a strong usage format can still be platform-specific.
Origin, Production & Authenticity
Estate provenance, harvest practices, certifications and traceability
This cluster rewards production evidence rather than retail familiarity alone. Castillo de Canena nearly matches Cobram Estate, while Oro Bailén and Gaea enter the top five through origin, estate and production associations that are less dominant elsewhere.
Retail, DTC & Gifting
Online purchase, subscriptions, gift sets and replenishment
The order changes sharply after California Olive Ranch. Brightland, Kosterina and Graza rise because their digital merchandising, gifting language and direct-purchase architecture give AI systems more explicit commercial evidence than traditional category prestige alone.
It ranks first in all five clusters and appears in every cluster-platform combination. Its lead is strongest in Cooking & Culinary Use, where it enters 157 of 240 responses. The brand is not relying on one premium cue; it is legible as healthy, flavorful, practical, traceable and purchasable.
Cobram Estate Olives’ US$173.5 million acquisition of California Olive Ranch created a portfolio containing California Olive Ranch, Cobram Estate and Lucini. They rank first, second and seventh, respectively, while retaining distinct consumer identities.
Brightland, Kosterina and Graza all rise into the top four behind California Olive Ranch. Their advantage is not explained by broad category heritage alone. AI systems can clearly connect these brands to online purchase, replenishment and gifting situations—the commercial layer where polished product pages and offer architecture matter.
Cobram Estate is Claude-heavy. Colavita is ChatGPT-heavy. Graza and Partanna are strongest on Gemini. Olivea is absent from Claude and ChatGPT. A single blended rank therefore cannot diagnose a brand’s actual visibility problem; the platform pattern determines what kind of evidence is missing.
Category-wide evidence. Brands become defaults when the web explains not only what they sell, but why they fit multiple buyer situations: cooking method, flavor, health criteria, provenance and purchase context.One strong association can win a cluster. It does not create category leadership.
Machine-readable commercial context. The DTC cluster rewards brands whose product pages, gift sets, subscriptions and purchase options make the next step explicit. Attractive branding without structured commercial evidence leaves the model to infer too much.AI systems need a reason to recommend the product and a clear path for the buyer to obtain it.
Cross-platform reinforcement. Trained-model familiarity, current web retrieval and source selection do not reward the same evidence equally. Brands with visibility on only one system have momentum, not durable inclusion.The goal is not a single high score. It is repeated relevance across different recommendation environments.
Turn positioning into answerable evidence. A claim such as “high polyphenol” or “ideal for finishing” needs a dedicated, crawlable explanation with definitions, measurements, harvest context and clear product applicability. Broad lifestyle copy does not give an AI system enough justification to choose one brand over another.
Publish for retrieval, not keyword repetition. KDD 2024 research found that adding credible citations, relevant quotations and statistics improved source visibility in generative-engine responses; keyword stuffing did not.2 For olive oil brands, that means substantiating production, freshness, sensory and health claims with specific, attributable evidence.
Close the platform-specific gap. A Gemini-heavy brand needs broader authority beyond current web retrieval. A ChatGPT-heavy legacy brand may need more current, structured DTC evidence. The corrective strategy should follow the platform pattern rather than applying the same content plan to every visibility problem.
AI recommendation systems are already constructing a premium olive oil shortlist before a buyer reaches a shelf or product page. The window is open now for brands to define the evidence those systems use.