On store shelves at Uniqlo, a Jeju tangerine-print T-shirt has been restocked ten times. Cumulative sales: 70,000 units — a number far removed from the usual collab merch that shines for one season and vanishes. The partnership wasn't led by one of Jeju's major agricultural cooperatives, nor by a decades-old distribution conglomerate. It was led by Medals, a startup helmed by CEO Yang Je-hyun.
One figure gives this scene its context. Korea's fruit-growing industry generates more than 7 trillion won (roughly $5 billion) in annual output. Yet there's no domestic fruit brand that consumers reflexively recall the way they do Zespri or Sunkist. That gap is where Medals started, with a question the founders asked themselves early on: "Why doesn't Korea have its own Zespri?" Production is plentiful. Memory is not. Gyuldal — the tangerine brand at the center of this story — began in that blank space.
From varietal codes to the language of taste
For a long time, Korea's tangerine market has been classified and sold by varietal. Greenhouse-grown or open-field, cheonhyehyang or hallabong — these were the operative categories. For farms and distributors, that made sense: it matched their internal management systems and simplified pricing.
The problem is that consumers don't actually think in those categories. Few people wake up thinking, "I feel like eating greenhouse tangerines today." At the produce shelf, the deciding factor is something else entirely — do I want something sweet, something tart, what mood do I want this fruit to give me? The decision happens in the register of taste, not taxonomy.
Gyuldal moved its classification system into that register. Sweetness, acidity, body — sensory vocabulary borrowed from wine and coffee, applied to tangerines. There's a decisive difference between "this one is high in sugar with soft acidity" and "this is a cheonhyehyang." The first connects to what I want right now; the second stays a catalog code. As CEO Yang Je-hyun put it: "For an industry to grow, it needs language that consumers actually remember."
Zespri, Dole, and Sunkist were all built in that same direction. What sticks in memory isn't the fact that a kiwi comes from New Zealand, but the specific impression of that green color paired with its tartness. "California sunshine" compresses the taste and aroma of an orange into two words of lived experience. Origin and varietal are producers' categories; what these brands built was language rooted in the consumer's experience.
Gyuldal took this one step further. Through taste-based positioning, it moved Jeju tangerines into the register of lifestyle. Not just onto the produce shelf, but into the daily life of consumers who already associate something with Jeju. The 70,000 Uniqlo collab T-shirts and ten restocks are the result of that shift. Medals is now extending the same playbook to other regional produce — peaches, apples, pears — and that same momentum carried it to pop-up stores in Los Angeles and on the Stanford campus.
Branding can't come before quality
A familiar objection comes up on the farming side: no matter how polished the packaging and story, a brand won't survive a single season if it can't reliably deliver fruit that meets its stated sweetness standard. In this view, a producer needs to lock down quality control before refining how it's positioned to consumers.
That objection has grounds. Korea has seen agricultural brands that grabbed early attention with clever storytelling and design, then disappeared within two or three seasons because they couldn't hold onto repeat customers. When expectations are built up but the product doesn't deliver, that gap turns into backlash — and the more precise the positioning, the more specific the disappointment.
This matters for reading Gyuldal's success correctly. Repeat purchases followed because Gyuldal actually built its sweetness-acidity-body classification into the real product lineup. Brand trust is built on consumers repeatedly experiencing, "they said this one would be low-acid, and it was." Closing the gap between description and reality was the precondition, not an afterthought.
But taking the objection at face value leads to its own trap. Even after quality is in place, a product goes unremembered if there's no taste-based language delivered to the consumer. Korea's fruit market already has plenty of farms producing excellent fruit. The reason none of them became a name consumers instantly recall isn't a quality problem — it's that positioning work beyond origin and varietal information has been rare. Quality and branding aren't sequential. Both are required, in parallel.
The café market already lived through this same parallel structure. Once bean quality and extraction technique rose across the board, some cafés stopped there, and others went one step further. That extra step was defining not just coffee quality, but what sense, routine, or mood the space and the cup connect to in a customer's life. A pattern shows up again and again among cafés that built real brands: the ones that defined their menu stay comparison items; the ones that defined an experience build regulars and fans. Read the stories of cafés that first decided what start-of-day feeling they were actually selling before they sold coffee, and Gyuldal's chosen direction stops looking like an eccentric choice for the agricultural sector — it looks like the same move, made in a different aisle.
Whose language is your positioning written in?
Reading Gyuldal's case as a niche success story in agricultural branding leaves less to learn from it. The structure that repeats here is broader than that.
Classify tangerines by varietal, and your competitors are other farms growing the same varietal. Classify them by sweetness, acidity, and body, and you occupy a different position on the map of consumer taste. "Tangerines you choose the way you'd choose a wine" was a space that didn't exist in the tangerine market before. Where you draw the category boundary determines who you're actually competing against.
That's a direct question for solo operators and content directors too. Look at how you currently describe your own service or product — is it written in the producer's classification, or in the language of the outcome the customer experiences?
"Email marketing consulting" is a producer's classification — it defines the work from the consultant's own vantage point. "The newsletter that got a customer to reply for the first time last month" is a description of the outcome the customer experiences. Both phrases point at the same service, but they don't carry the same power to move a decision. The first sits on a shelf next to dozens of similar services; the second lodges in the memory of someone who wants that exact result.
I don't think this is a copywriting problem — it's a positioning decision. The order matters: first decide which category you're competing in and which moment of the customer's decision you're stepping into, then translate that into language. Before Gyuldal ever devised sweetness, acidity, and body as categories, there must have been a prior positioning decision — "we're going to intervene in the moment of taste selection, not sit on the produce shelf."
There's a practical way to find that shift. Collect the words your customers naturally use after experiencing your service or product — unsolicited reviews, phrases that slip out mid-conversation, the words they reach for when explaining why they came back. The language customers use on their own is almost always more specific and more sensory than the categories producers define. That's where redefining your category starts.
In a 7-trillion-won fruit market, a gap in branding became one startup's starting point. In the market you work in, which spaces are still waiting to be filled by a consumer's senses?
