nephew: A Yoyogi-Park Café and Bar Discloses ¥39.96M in Its First Nine Months — 101 Guests at ¥1,294 by Day, 37 at ¥3,170 by Night

Real numbers from restaurants rarely surface. When they do, it’s usually a single line, “we do X a month”, with no breakdown. nephew, a café and street bar on a back street two minutes from Yoyogi-Koen station in Tokyo, is an exception. In January 2022, Takushi Izawa, who heads the studio & Supply that operates it, published the final installment of his opening diary with the full set: sales, profit, guest counts, average spend, and the gap between planned and actual startup costs.

The figures: from opening on April 20, 2021 through December, nine months produced ¥39,968,275 in sales and ¥7,841,308 in profit. Excluding April and May, when a state of emergency made bar service impossible, the seven months of full operation from June to December averaged ¥5,709,754 in monthly sales and ¥1,287,027 in monthly profit.

The original business plan called for ¥2,665,000 a month. Actual results more than doubled it. But the shape of that overshoot is not what you’d expect: day and night diverged from plan in completely different directions.

Day and night, broken out (June–December 2021)

The café accounts for 54.6% of the total. In the plan, the café was pencilled in at ¥615,000 and the bar at ¥2,050,000. The night was supposed to be the earner. Reality inverted that. The café came in at roughly five times plan. The bar at about 1.2 times. In other words, the “double the plan” result was produced almost entirely by the daytime side.

Structurally, though, the more important fact is that the two are nearly level. Guest counts differ by a factor of 2.7 (101 vs. 37), average spend by a factor of 2.4 (¥1,294 vs. ¥3,170). The two gaps cancel out, and revenue lands in the same place. Same rent, same kitchen, same interior, two completely different P&Ls run inside a single day. That is the design of this shop.

The third pillar is easy to overlook: original merchandise, averaging ¥155,000 a month and exceeding ¥300,000 in December. Small in absolute terms, but food and drink sales are capped by seat count and opening hours, while merchandise only needs stock and a shelf to lift the ticket. At 2.7% of ¥5.71M in monthly sales it looks marginal. Yet it covers nearly 40% of the ¥407,000 rent without adding a single seat. Combined with food making up 20% of bar sales, the shop is built to collect from several different line items per visit.

Work back from fixed costs and the plan was never going to be comfortable

The property is a two-storey detached house of 71.19㎡ (21.53 tsubo), with monthly rent of ¥407,000, or ¥18,903 per tsubo. Divide that rent by sales and the ratio becomes visible.

Against the actual monthly average of ¥5,709,754, rent runs about 7%, quite light for a restaurant. Against the planned ¥2,665,000, the same rent would be about 15%. At that level, very little survives after cost of goods and labour.

Put differently, this shop was in a position where landing exactly on plan would have left it immobilised by rent. Doubling the plan was very likely not a bonus but the condition under which the business works at all. Compare the crêpe shop in five tsubo that reached ¥1M a month and still couldn’t continue or the coffee stand that closed five months after being acquired: how far revenue stacks above fixed cost is what separates a surviving physical shop from a closing one.

Startup cost deserves a look too. Planned at ¥19M, actual came in at ¥28.5M, an overrun of ¥9.5M, or 50%. Divided by the ¥1,287,027 average monthly profit, simple payback runs about 22 months. Had the build come in on plan at ¥19M, it would have been about 15 months. Seven months of difference. The precision of the pre-opening estimate shows up directly as payback period.

What didn’t work — and what got switched off

The hardest thing in this case wasn’t a tactical mistake but an external one. The shop opened on April 20, 2021. Four days later a state of emergency was declared and bar service became impossible. April and May ran as a reduced mocktail bar closing at 20:00, with full operation resuming only in June.

Seen from there, the two-shift design takes on a different meaning. When the night was shut off by regulation, the daytime café could keep trading. The segment that had been cast as the lead earner disappeared entirely, and the supporting act carried the period. That said, this is a reading of the disclosed figures. The operator does not explicitly claim the day/night split functioned as risk diversification.

One more thing he does name as a success factor is an unexpected line: that he and his partners are not standing in the shop. The business is designed to run on staff (three on weekdays, four on weekends) without the owners working service. In owner-operated restaurants, the owner’s own hours are frequently absent from the cost line while quietly inflating profit. nephew’s roughly 20% profit-to-sales ratio over nine months is a figure produced without that hidden subsidy. The flip side: if the owner isn’t there, training and delegation have to be in place before opening, not after.

Nine months of a second venue, not a cold start

This was not a launch from zero. The location is a back street two minutes from Yoyogi-Koen station, in a detached house the business could occupy entirely. & Supply is a creative studio, and nephew is its second venue. It launched carrying the track record and brand of the first shop in Ikejiri-Ohashi, and by the nine-month mark its Instagram account had 9,400 followers. Monthly revenue at twice the plan has to be read with that head start included.

The figures need care as well. Everything cited here is from 2021. The shop is still trading. The official venue page lists the café at 8:00–17:00 and the bar at 19:00–24:00, shifted from the 10:00–17:00 and 19:00–01:00 recorded in the opening diary. Recent sales have not been published. Physical shops sometimes look best in their opening period and sometimes take far longer to ramp. What this article covers is one cross-section: the first nine months.