I went in with expectations. Seoridan, in Mulgeum, Yangsan, was featured last summer on TV Chosun's 〈식객 허영만의 백반기행〉 (Huh Young-man's Baekban Journey, a TV Chosun dining show) as Yangsan's specialist in memil (buckwheat) noodles and ongsimi (potato dumplings)1, and it has even earned a Blue Ribbon Survey listing2. The modern hanok, converted from an old post office building, looked immaculate even in the photos3, and sure enough, the place was nearly full when I arrived. Up to that point, there was nothing to fault.
The trouble began when the food arrived. The buzz I had caught from the atmosphere settled at the first bite, turned to puzzlement at the second dish, and finally hardened into a single question. With this many people here, why don't I want to come back?
As someone who runs a restaurant myself, I don't go around bad-mouthing places whose food falls short. But a thriving restaurant is always doing something well — and when that something isn't the food, the gap between the two teaches you more than anything else.
A beautiful bowl brings a guest once. What brings them back is always what's in it.
Time on the Plate
We started with the perilla-oil memil (buckwheat) noodles (12,000 won)4, billed as the signature dish. To be honest, the perilla-oil buckwheat noodles they sell at Costco tasted better. The noodles were fine, the perilla-oil aroma was fine, and fine was as far as it went. It never took a single step past a flavor you've had somewhere before. The 12,000-won price kept nagging — not at my palate, but at the back of my mind.
The black-sesame ongsimi (potato dumplings) (13,000 won)4 my wife chose earned a verdict of "just okay." The nuttiness of the black sesame was there, she said, but the broth had no umami underneath to support it, so the flat first impression of the opening spoonful carried through, unchanged, to the last. Nuttiness is aroma; umami is structure. Aroma without structure wears you out before the bowl is empty.
The herring-roe pork-jowl suyuk we ordered on the side was the biggest disappointment. Pork jowl is, by nature, a cut densely marbled with fat. It only turns properly savory when it's grilled over fire and the fat renders off — but boil it into suyuk and the fat, with nowhere to go, sits squarely on the tongue. A few slices in, we'd had enough. It was a dish where the cut and the cooking method were at odds — a plate that should have been culled at the planning stage.
The thread running through all three plates was clear. Nothing was inedible, but nothing was worth coming back for either. And that, precisely, was the starting point of the question that followed me out the door.
Why the Line Forms Anyway
With food at this level, why is the place doing so well? Taking the restaurant apart in my head in search of an answer, I found it engineered far too precisely to be lumped under the label "Instagram restaurant." The machinery for drawing in customers wasn't one layer deep — it was several.
First, media. The Baekban Journey broadcast and the Blue Ribbon12 are, in themselves, an enormous acquisition funnel. Second, space and story. The narrative of a hanok converted from an old post office3, plus details like a restroom stocked with air conditioning, mouthwash, dental floss, even a diffuser5, function as "content points" that make guests want to photograph and post. Third, the district. Seoridan doesn't stand alone; it sits inside "Seoridan-gil," an alley clustered with cafes, a bookshop, and an Italian place6, so the whole neighborhood becomes a day-out course. Fourth, reputation management. Volunteer work — hosting free meals for grandparent-headed families and the elderly7 — makes the local paper and lays down a foundation of goodwill.
But what struck me most was the structure of the reviews. At the time of my visit, Naver showed roughly 24,000 visitor reviews against barely 800 blog reviews — a gap of about 30 to 18. The difference between blog posts, written voluntarily, and visitor reviews, rewarded with a freebie like memil dumplings for snapping a photo of your receipt5, is engraved right into that ratio. The figure of 24,000 measures not the sum of satisfaction, but the sum of a well-designed review-generation machine.
Acquisition brings a guest to the door. Food brings that guest back next month. They are entirely different engines.
The Two Engines That Drive a Restaurant
A restaurant runs on two engines. One is the acquisition engine — the power to pull a stranger through the door. Broadcasts, interiors, reviews, the district, marketing all belong here. The other is the return engine — the power to make someone who came once come again. That one rests almost entirely on the food, and on the food's ability to convince you that the money was well spent.
Seoridan has built an acquisition engine that ranks among the best in Yangsan. The problem is that its return engine is all but switched off. Strong acquisition keeps you going for a while — the broadcast effect and a stream of first-timers keep filling the empty seats. But if those guests never come back a second time, the restaurant is locked into a structure where it must keep buying new customers forever. Pour as much water into the top of the funnel as you like; if there's a hole in the bottom, the bucket never fills.
Let there be no misunderstanding. Building strong acquisition is genuine skill. The idea that good food alone will bring customers is a romantic delusion that kitchens with good food fall into all the time. But acquisition and return are bound by multiplication, not addition. When one side is near zero, the result converges to zero no matter how large the other grows.
How to Read a Thriving Restaurant
Envying a restaurant that's doing well is easy. The hard part is dissecting exactly 'what' it is doing well on. Seoridan's success is not a success of food; it is a success of acquisition design. Blur that distinction and you end up copying the wrong things.
What is worth learning from this place is clear: the mechanism that nudges receipt reviews along naturally, the details that make guests want to leave with photos, the instinct to set up shop not as a lone outpost but along the walking path of an alley district. All of it widens the top of the funnel without touching the substance of the food.
What to guard against is even sharper: a plate where the cut and the cooking method disagree, a broth with aroma but no structure, a bowl that calls the price to mind before the flavor. It is the moment when confidence in a well-built acquisition engine slides into the delusion that the return engine can be left switched off.
In the end, a good restaurant is one where both engines run together: the power to bring someone once, and the power to bring them back. Seoridan was the clearest possible textbook for seeing those two forces separately. Though I doubt I'll ever be going back.