More and more shoppers are skipping the search bar and simply asking an AI assistant what to buy. So we asked five of them the same 45 Turkish-language questions five times each about pet food, litter, grooming and equipment in Turkey, producing 1,124 analysed responses. Here’s what they named, where they disagreed, and how asking specifically for Turkish or ‘yerli’ brands changes the visible market.
Type “en iyi kedi kumu hangisi” into ChatGPT and you’ll get a confident, friendly answer in seconds. Ask Gemini the same thing and you’ll get another confident answer, possibly with a different list. Ask Grok and you may get nine or ten names, starting with a marketplace rather than a brand.
This is the quiet shift happening in how people find products. Discovery is moving from the search engine, where you scroll a page of links and decide for yourself, to the AI assistant, where a single answer arrives pre-sorted. When the assistant names three products, brands or places to buy them, those names have been placed on a shelf at eye level and everything else is in the back of the store.
Why this is worth paying attention to
Turkey is an unusually cat-oriented pet market, and an unusually fast-growing one. KONDA’s July 2024 national barometer found 13% of respondents had a cat at home and 8% a dog, while the European pet food federation FEDIAF estimated Turkey’s owned-pet population at roughly 4.9 million cats against 1.5 million dogs in 2024. The Ministry of Agriculture and Forestry’s own PETVET registration figures, reported in July 2024, showed 1.42 million registered cats against 950,125 dogs. The three measures produce different ratios, but they all point the same way.
The commercial side has grown quickly. The number of businesses holding ministry permission to manufacture cat and dog food rose from 46 in 2020 to 114 in 2024, and the ministry’s 2025 list carries 106 approved production establishments. Exports of retail cat and dog food reached US$145.2 million on 97,992 tonnes in 2024 on ministry figures using customs data, and Aegean exporters’ association data put 2025 at around US$150 million across 112 countries. The Ministry of Trade’s e-commerce monitoring reported that marketplace turnover in the “Evcil Hayvan Ürünleri” category was 49.3% higher in November 2025 than a year earlier.
There is no official national figure for the size of the retail market, and the published estimates disagree sharply enough that we won’t pick one. What is clear is the direction: more animals in more homes, more domestic factories, and a category moving online fast.
Meanwhile, the channel we’re measuring here has arrived at scale. TÜİK’s first official measurement of the subject found 19.2% of internet users aged 16–74 used generative AI in 2025, while GWI’s survey work put 39.7% of Turkish internet users on ChatGPT in the previous month in Q2 2025 — different definitions, both describing something real. Globally, GWI found 22.1% of online adults using AI tools to research brands by late 2025, against 45.8% still using search engines, with the search figure falling from 51.3% eighteen months earlier.
All of which makes one question worth asking directly: when someone types a Turkish question about cat food into an assistant, whose name comes back? Over a single day in late August 2026 we put the same 45 Turkish-language questions to five leading assistants — Claude, Gemini, ChatGPT, Perplexity and Grok — asking each question five times. That produced 1,124 analysed responses and 5,713 tracked-entity mentions.
One brand led, but it didn’t lead everywhere
The clearest single result was Reflex. It appeared in 25.53% of all 1,124 answers — 287 of them — and was named by all five assistants, putting it top of our overall leaderboard at 87.92.
Below it the list thinned fast. Of 225 tracked entities — brands, sub-brands, retailers and marketplaces — only 28 cleared the 5% mention threshold in the general cut. Across all ten cuts, 82 qualified at least once. Visibility is not evenly distributed: a small group of names occupies most of the space and a very long tail barely registers.
There is a useful piece of context for the leader. Reflex is made by Lider Petfood, which said in March 2026 that it holds close to 40% of Turkey’s domestic dry pet-food market — a company estimate rather than an audited figure, but a plausible explanation for why one name dominates the answers. The same group’s litter brand, Proline, is second in our litter and grooming ranking.
The current ranking: the top 28 entities by AI visibility
Below are the 28 entities that cleared the 5% mention threshold in the general cut, ranked by AI Visibility Score. The list includes brands as well as marketplaces where the assistants named them as answers. This is a visibility ranking, not a quality ranking.
For all 225 tracked entities, the methodology and the category breakdowns, see the current edition below.
Asking for yerli brands changes the visible market
This is the finding we didn’t expect. Twenty-two questions asked explicitly for Turkish or yerli brands; the other 23 said nothing about origin. The two groups produce strikingly different rankings.
On the questions that ask for Turkish brands, not a single foreign entity clears the threshold. Of the 25 qualifiers, 24 are verified Turkish and one is of unknown origin; Turkish entities account for 98.39% of qualifier mentions.
On the questions that don’t mention origin, the entire top seven is foreign — Royal Canin, Farmina, Acana, Science Plan, Ever Clean, Furminator, Orijen — and Reflex, the study’s overall leader, drops to eighth. Foreign entities account for 66.21% of qualifier mentions.
Look at what those foreign names have in common and the pattern gets sharper. Most are premium imported dry food or specialist litter and grooming brands. Turkish entities hold much of the everyday shelf: Turkish brands take the top two places in litter and grooming, while Turkish entities take the top three in equipment and accessories — including Trendyol, a marketplace, at #2. The divide isn’t simply about nationality; it is also about which part of the aisle you’re standing in, and imported premium nutrition is where domestic visibility thins.
That maps onto something real in the market. Turkey has substantial domestic manufacturing capacity, with published plant capacities of 220,000 tonnes a year at Lider, 120,000 at Hermos and 45,000 at Çağatay. But imported premium food has held a structurally important position for years: a 2021 development-agency study using official data put imports at 44% of pet-food sales value in 2019, and a US Department of Agriculture report in February 2024 found a 15 kg imported premium bag selling at around US$69 against roughly US$30 for a locally produced one.
The assistants don’t see the same market
You might assume five AI systems trained on broadly similar data would converge. They didn’t. Grok produced 10.02 tracked-entity mentions per answer against ChatGPT’s 2.42 — a four-fold difference — and Grok names shops heavily, with Trendyol appearing in 72% of its replies, ahead of any actual brand.
Perplexity read far more of the web before answering than anyone else, citing 11.87 domains per answer against ChatGPT’s 2.81. Reflex came first for three of the five assistants; Claude put Royal Canin first instead.
“Web search” and “memory” produce different winners
Not every AI answer is built the same way. Sometimes the assistant searches the live web; sometimes it answers from what it already knows. Four of our five searched on essentially every answer. Gemini searched on only 43.3% of its own.
That gives a rare side-by-side. When Gemini searched, it named Turkish brands — Xenon Smart, Catrevi, ATA HOME and others turned up regularly. When it answered from memory, those brands appeared not once, and their places were taken by Farmina, Royal Canin, Acana and Orijen.
One honest caveat, which we make at length in the report: Gemini also chose to search far more often on the questions that asked for Turkish brands than on the ones that didn’t, so we can’t cleanly separate the effect of searching from the effect of the question. What we can say is that on this evidence, established international brands are already inside the model’s memory, and smaller local ones have to be found on the live web.
Where the assistants learn about brands
Not from the brands. Retailer websites account for 33.04% of all the sources cited, against 15.49% for brands’ own sites. The single most-cited domain in the study was the pet retailer petlebi.com, cited by 280 responses — nearly a quarter of everything we collected. Its community subdomain, sosyal.petlebi.com, was cited by 103 responses; those response sets may overlap.
That fits how Turks actually buy. Sector data compiled for a March 2026 industry report put specialist pet shops at 28.2% of market value, supermarkets at 27.3%, e-commerce at 26.2% and veterinary clinics at 17.2%, while a February 2026 survey of 702 Turkish pet owners found 49% now usually buy online, up from 38% two years earlier. Several of the biggest specialist retailers have also become brand owners themselves — Petlebi has Rimba and Vet’s Plus, Petpal has WonderCat and WonderDog, Markamama has Obivan and Herniks — and some of those own-label names surface in the assistants’ answers alongside the manufacturers’.
The practical consequence for a brand is uncomfortable: how a product is described on a shop’s category page may matter more to its AI visibility than anything on its own website.
The most important finding isn’t a ranking
Here’s the thing we’d most like readers to take away: being recommended by AI does not mean a brand is good. When an assistant names a brand often, it usually means one thing — there is a lot of information available about that brand. It does not mean the food is nutritionally better, the litter clumps better, or the product is right for a particular animal.
That matters more in this category than most. Turkish pet owners still overwhelmingly trust their vet: a 2025 survey of 519 owners found 42.58% cited veterinarian recommendation as their influence on pet food choice, against 9.83% for anything online. An AI assistant is not a vet, and nothing in this report is veterinary advice. AI visibility measures how easily a brand is found, not how good it is.
Read the full report
This article is a summary. The complete report contains the full 28-entity general leaderboard, the breakdown for each AI assistant, the segment-by-segment winners, the source analysis, the search-versus-memory comparison and the full methodology. This is the first edition of a study we’ll repeat every quarter; you’ll find the current and past editions listed on the right side.
Frequently asked questions
Reflex leads with an 87.92 score, appearing in 25.53% of all 1,124 answers and named by all five assistants. Royal Canin and Felicia complete the top three. Different parts of the market have different leaders: Royal Canin in pet food, Xenon Smart in equipment and accessories, and VanCat in litter and grooming.
It depends entirely on how you ask. On questions that specifically request Turkish or yerli brands, 24 of the 25 qualifiers are verified Turkish, one is of unknown origin, and no foreign entity qualifies. On origin-neutral questions, the top seven qualifiers are all foreign and foreign entities account for 66.21% of qualifier mentions.
Royal Canin is the strongest, topping the origin-neutral ranking with the highest position score in the study. Farmina, Acana, Science Plan and Orijen follow, all of them dry food. Outside food, Ever Clean and Sanicat appear in litter, Furminator in grooming, and KONG, Trixie, Ferplast and Xiaomi in equipment.
On the leaders, largely yes — Reflex is the most-named brand for three of the five. On everything else they differ sharply. Grok produces 10.02 tracked-entity mentions per answer against ChatGPT's 2.42, and Grok often answers with a marketplace rather than a brand, naming Trendyol in 72% of its replies.
No. A high AI visibility score means these systems have a lot of information available about a brand and surface it readily. It does not measure nutrition, safety, quality or suitability for a particular animal. Visibility is about how easily a brand is found, not how good it is, and nothing here is veterinary advice.
Claude Haiku 4.5, Gemini 3.6 Flash, ChatGPT 5.6 Luna Pro, Perplexity Sonar and Grok 4.3 were each asked the same 45 Turkish-language questions five times, on 28 August 2026, producing 1,124 analysed answers.
About Herm.io & disclosure
Herm.io is a consumer behaviour and marketing data company. We study how people discover and choose brands to help businesses reach the right customers. This report forms part of our public research and is conducted quarterly.
Disclosure & neutrality
The rankings and metrics in this report are entirely impartial and are based on the methodology described within the report. No brand can pay or provide sponsorship to be included in the report, improve its ranking or influence how it is described in the content.
Mert Can Elkaya
Contributor
I'm a product builder working at the intersection of product, fintech, and growth. From martech and venture capital to leading product at a proptech platform and co-founding a fintech startup, I help teams—and shoppers—make smarter, more confident decisions.
More from Mert