369 Cups and a 15-Kilogram Backpack: The Real Balance Sheet of NPB's Beer-Vending Model

In the first week of October 2026 a video from Yokohama circulated: a spectator bought a stack of draft beers in one order and poured them into the discarded-drinks bin. The same week The Japan Times published an interview with Mana Kato — 21 years old, three seasons walking the stands at Belluna Dome for the Seibu Lions, 100 cups on her first day, a personal best of 369. The two numbers belong to the same model. 100 cups is the productivity the recruitment content sells; 369 is the productivity the concession revenue is built on. The 3.7× gap between them is not a talent story. It is a piece-rate cost structure.

Define the terms first, because the rest of the argument rests on them. NPB's mobile draft service is a service design, not a tradition. MLB parks sell chilled cans and bottles from the aisle — transactional, low-labour, low differentiation. In Japan the vendor walks to the seat, pours from the tap, and remembers the regular's name. Per The Japan Times, that system has become a signature of NPB stadiums.

The labour supply is just as specific. Kato applied in 2026 after watching TikTok videos, knowing almost nothing about baseball. Kuala Lumpur taught me that fandom travels faster than the official press release — and here the recruitment pipeline runs on lifestyle content, not on fandom. The tenure ceiling is explicit: this is a student workforce, and graduation ends the job automatically. Her October 2026 exit is that mechanism, not an accident attached to the viral clip.

The valuation habit came to me in Kuala Lumpur while building a football-sponsorship model for a telco — cost first, story second. Here the cost is plain: a pack of roughly 15 kilograms, three to five hours of continuous stair traversal, humid heat, short water breaks in between. That is not a description of a stadium experience; it is a manual-handling job description. In 2026, during the FIFA U-17 World Cup in Guwahati, I built a time-stamped signage audit across 52 matches, 41 pages, per-venue exposure — and even there the picture on the ground and the picture in the file were two different documents. Same gap here.

I found the real story in the eight-to-one footnote — in this case, the 100-to-369 footnote. A first day at 100 cups and a peak at 369 three seasons later tells you the job is not low-skill; it has a steep learning curve and a high skill ceiling. In a piece-rate system that much dispersion usually produces two outcomes: novices exit within a season, and a small, high-earning veteran core remains. Kato was in that core — she recruited her younger sister, which means part of the hiring runs on referral, not a formal HR channel.

The income arithmetic rests on terms that are not on the record: pay depends partly on cups sold. Per-cup commission, base wage and employment classification — employee or contractor — are all absent from the report. No rate, no model, so I will not build one; I will only mark the direction. If a 15-kilogram pack carries 10 to 15 litres of net beer and a pour is 350–400 ml, then 369 cups implies roughly 10 to 15 reload cycles in a game. That is a derived estimate, not a number in the article — keep confidence low.

The most important business fact is whose name the customer relationship sits under. The regular-customer base is the engine behind 369 — the vendor introduces herself, recalls the name, the customer returns. That capital belongs to the individual, not the club. The day she graduates and leaves, the club loses a revenue-generating relationship it never booked. The spreadsheet was right; the market was just slower to admit it.

Reading Yokohama as a fan-behaviour story is the easy read and the wrong one. The day the incentive design and the boundary question become the same question, that is the model's real blind spot surfacing. Piece-rate pay and relationship-based selling together hand a spectator a legitimate reason to buy ten cups he will not drink. The cups do not hit the bin because the fan is bad; they hit the bin because the real product in that transaction was attention, not beer.

The second failure is documentary. The report carries no operator statement, no quoted rule text, no sanction. Yuichi Ito, 56, a long-term fan, says he does not secretly photograph staff and does not follow them into corridors — which tells you that it happens. What has no written rule is left to self-policing. Kato's line that she was lucky never to have encountered a frightening situation is a risk-management red flag, not reassurance: it credits safety to chance rather than to a system.

Third, the arithmetic does not close. Applied in 2026, three-year tenure, quit in October 2026 — that reads closer to four years. Verify the publication date and the tenure count before citing any of this externally.

The number that matters is the one no file contains: the base rate of harassment, vendor turnover, and the per-cup rate. The 12 NPB clubs govern club by club, so the next action will arrive in fragments. The question is whether by the 2027 season operators pair a name-recall service with a written protection channel — or repeat the same absence until the next clip.