¥1M a Month in 6 Months, 13 Months and 14 Years: the Fastest Cannot Step Away from the Work

We read down the frontmatter of every published case and pulled out the ones whose monthly-revenue field reads exactly 1,000,000. Of the 385 cases live on 13 September 2026, seven do. All seven are Japanese, all are still operating, all are run by one person or as a side business. None of that was a surprise.

What stopped us was the next field over, the time those seven took to reach ¥1M a month. Even looking only at the cases whose period can be pinned down from the operator’s own account, the short end is six months from opening and the long end is 14 years. Same country, same size of operating team, same number coming out the far end. The time to get there differs by a factor of roughly 28. And once they are sorted by that time, the operators who arrived fastest turn out to be the ones who can no longer pull themselves out of the business.

Three of the seven go into one frame here: what they sell, how they set the price, and what happens to revenue when their hands stop. The criteria we used, and the reasons the other four are out, are disclosed at the end.

Six months, 13 months, 14 years

The amount is identical across all three, and almost nothing except the amount lines up. We take them one at a time.

Three ways to sell, but only one pair of hands at the griddle

In November 2021 a 23-year-old rented a 5-tsubo kitchen in Osaki, Shinagawa, and opened a crepe shop. Solo operation, five days a week, six hours of service a day, closer to eight once prep and cleanup are counted. December, the first month after opening, brought roughly ¥500,000 across the three channels of counter, delivery and mail order. By April and May of the following year there were months above ¥1,000,000. Among the cases whose period can be identified, that is the fastest arrival.

The design itself is sound. A counter on its own caps revenue at the number of people walking past five tsubo of frontage. Delivery widens the catchment to a radius measured in kilometres, and mail order takes geography out of the calculation. The build separates the revenue ceiling from the location, and from the first month it never rested on a single channel.

Eight months in, the owner wrote that the business had been a failure. It was profitable. The startup costs had been recovered. The reasons the owner gave were that the owner personally is the shop’s concept and therefore cannot leave the shop, and that solo operation left nobody to share the events of a shift with.

Break the structure down and it becomes a plain story. There are three ways to sell, and in every one of them the person cooking the crepe is the owner. Revenue grows by adding hours or by raising the price, and the route of hiring people to widen the operation stays shut for as long as the owner is the concept. The more revenue grows, the stronger the reason the owner cannot leave the floor. The personal quality that served as a weapon during launch converts straight into a cost at the moment of expansion or exit.

We do not read that “failure” as a retreat. What the owner logged is the position of the ceiling, found six months after opening.

The price rose 7x, and the person delivering is still the same one

PAPANAVI, a designer of 18 years, took zero jobs in the first six months after signing up to Coconala. Profile filled in, portfolio posted, and still no requests at all. The operator’s explanation for the silence: “there are mountains of services like that on Coconala.”

Two decisions came out of it. The scope narrowed to beauty salons and makers of beauty equipment, and the floor price for a logo went from ¥3,000 to over ¥20,000, roughly seven times higher. In month ten after going independent the listing ranked first among 39,028 in the design category, and about a month later monthly income reached around ¥1,000,000.

Those two are faces of a single operation rather than separate tactics. A generic logo listing is one item among tens of thousands, and the only axes left for comparison are price and review count. The moment the listing says “for beauty salons,” the comparison set switches to the narrow pool a salon owner searches through. Sitting inside the same category, the designer cut the effective number of competitors. The price rise does the work of replacing the comparison axis itself, and the operator moved it after observing a buyer-side reaction: too cheap reads as low quality.

Even at seven times the price, the one delivering the file is the designer. About 60% of revenue comes from repeat clients, and the structure in which a logo is followed by business cards, shop cards and flyers stacks demand on top of demand. It is a strong structure, and all of it runs through the working hours of one person. The difference from the crepe shop is where the work happens. Revenue wired directly to the operator’s own hours is the same in both.

220,000 fewer pageviews, ¥200,000 more income

The third case runs on a different order of magnitude of time. SiRO, who spent about ten years at a systems integrator before going full time, needed 14 years to reach ¥1,000,000 a month. Nothing outsourced. Around 15 servers and ten domains come to a combined ¥20,000 to ¥30,000 a month in fixed costs.

What makes this record valuable as a document is that every income step is published paired with the pageview count at that moment. The last two steps have a strange shape. At ¥800,000 a month the sites did 4.44 million pageviews. At ¥1,000,000 they did 4.22 million. Pageviews fell by 220,000 and monthly income rose by ¥200,000.

What was happening in the later years was therefore no race to add pageviews. Revenue per pageview had flattened out somewhere around 2.67 million pageviews, and pulling it back up from there is what ¥1M a month actually consists of.

The first five of those 14 years went in a loop of building and scrapping, and the cause was a bet on user-submission services. An unknown individual has no pull to gather submissions. From there the policy switched to “stop making submission-type services and build useful tools instead.” In a submission model the source of value depends on other people’s behaviour, and with zero users the value is zero as well. A tool carries value in its own function from the moment it ships, and a visitor arriving from search gets their business done even while the thing sits unattended.

This is where the distance between revenue and the operator opens up for good. A crepe earns nothing until it is cooked, and a logo cannot be invoiced until it is delivered. A tool keeps running while its maker sleeps. Reading those 14 years as the price paid to build that distance is the straightforward reading.

All three moved a unit price. What differed is whether their own time sits inside that unit

Something came into view only once the three were lined up. The variable each of them moved just before ¥1M a month was the price attached to one unit, and never the quantity of units.

The crepe shop took its ways to sell from one to three, the Coconala designer multiplied the unit price by seven, and the indie developer restored revenue per pageview. Nobody reached ¥1,000,000 by working more hours. That much the three share.

What splits them comes after: whether the operator’s own time sits inside the unit that price covers. One crepe contains the owner’s time in full. One logo contains it too. One pageview contains none of it. This single point explains the gap in time to arrival and the gap in freedom afterwards at the same stroke. Sort them fastest first and the order matches sorting by how much of the revenue is made of the operator’s own hours. The owner who arrived in six months cannot step off the floor, and the developer who took 14 years came away holding an asset that turns over on ¥20,000 to ¥30,000 a month.

This next part is speculation, but the order is no accident. A format that spends your own time starts fast, and in exchange every rise in revenue is a rise in the hours you are tied down. A format that spends no time returns almost nothing for the first several years. SiRO’s five years of failures were no reflection of personal ability. We read them as the way this format starts by its nature.

¥1M a month is a fork where the difference in format comes to the surface, not a destination. Three people reached the same number and are carrying completely different problems into the next year.

How we picked these three, and the four we left out

We did the picking, so here are the criteria. Of the seven, these three satisfied both of the following: the time taken to reach ¥1M a month can be pinned down from the operator’s own account, and the relationship between revenue and the operator’s hours is explained in the operator’s own words.

The four we left out have their reasons too. Tsuzuki Blog is already covered in the five owned-media cases. manablog and Unico Blog report figures combined across several sites, and the handmade-jewellery case is in its ninth year with wide month-to-month swings. None of the four could be placed on this column’s axis, the distance between one person’s hours and the revenue. Since ease of fitting that axis is what the exclusion rests on, the three that remain lean toward supporting it.

There is one more limit. The seven are the seven who felt like publishing a month at ¥1,000,000. Operators who stalled at the same figure, and operators who fell back from it, never entered the population. And this is revenue, not profit. The crepe shop carries cost of goods and rent, the design work carries production hours, and ¥20,000 to ¥30,000 a month of fixed costs against ad income is a different premise entirely. The single point that lets the three stand side by side is the top line, and with n=3 we publish no median.

For the distribution, the aggregate of monthly revenue across solo and side-business cases is in how far one person can go, and medians by business type are in the business-type map. Definitions of the metrics live in “how to read the numbers” on the data page.

The same amount does not make two businesses comparable

Having finished lining the three up, our trust in monthly revenue as a unit has dropped a little. The same ¥1,000,000 is, for the shop owner, ¥1M bundled with five days a week and eight real hours a day of being tied to the counter. For the developer it is ¥1M of which nearly everything stays in hand once ¥20,000 to ¥30,000 of fixed costs comes off. Placing those two on the same row and comparing which is larger carries very little meaning.

We built the case list that sorts by amount ourselves, and we will say it anyway: that ordering expresses nothing about which business is better. What it shows is the distribution of amounts, never the distribution of how many hours went into them. Of these three, the crepe shop was the only one that left a record of hours worked. That is what we want to count next time.