Company history

Seria began in 1985 as one man with a stall. Kawai Hiromitsu sold household sundries in the event spaces that Japanese supermarkets rent out by the week — a trade with almost no fixed cost, and almost no future. A rotating pitch gives you no stable catchment, and without a catchment you cannot plan a range or design a supply chain. Kawai spent the next decade converting the stall into a company: incorporation as Sanyo Agency in 1987, a head office and distribution centre built in Ogaki, Gifu, in 1988, and in 1994 the first permanent store, inside a Nagasakiya department store in Gifu.

He never disowned the origin — “the ¥100 shop owes its upbringing to the supermarket event floor,” he wrote later — and the habits of the itinerant trader carried into the fixed store: read the customer in front of you, and turn the goods fast. What the permanent store added was continuity, and with it the ability to manage an assortment rather than a load. From the start Kawai described the concept as a stylish variety store, competing on design and quality. That choice mattered because the alternative was hopeless: the industry was dominated by Daiso, which ran some 40,000 SKUs with private label at roughly 80% of them. Seria could not win a contest of scale, so it declined to enter one.

In 1997 Kawai committed to building an in-house ordering system — a decision that ran directly against what the industry believed. The received wisdom in ¥100 retail was that sales followed the number of items on the shelf, and that POS and ordering systems could never repay their cost on a ¥100 ticket. Kawai read the arithmetic the other way. In a format where you cannot move price by a single yen, every unsold item is dead inventory, and the whole result turns on the hit rate of what goes onto a finite shelf. Human memory could track that for a few dozen stores; it could not track it for hundreds.

The bill came first and the benefit later. “I ended up with debts I had not expected — it wore me down financially and mentally,” he told an interviewer in 2000. He compounded the difficulty by refusing the obvious cost lever as well: rather than buying cheap in volume, he told the trade press in 2001 that the rising cost of better own-brand goods would be solved “by deepening our partnership with the makers,” working small lots and managing cost jointly. Both choices pointed the same way — away from size, toward control. By $53.5M (¥7bn) in sales across 238 stores, Seria was a distant third in its industry, and the machinery that would change that was only just being switched on. In 2003 the company took the name Seria.

In September 2004 Seria put real-time POS into all 509 directly operated stores at once — a first for the industry, and one its own vendor resisted. To hold the cost to roughly half the normal figure, the project dropped the in-store controller entirely and had the registers talk straight to the head-office server, so that headquarters saw a sale at the moment it was rung up. The market was unimpressed. In June 2005, with rivals posting record profits, Seria’s earnings growth was expected to stall at 6%, the depreciation on the new system named as the reason; the view that a ¥100 business simply did not need this much management persisted in the trade for years. A JASDAQ listing in December 2004 funded the build.

The second half of the system arrived in 2006. Kawai Eiji — the founder’s successor, who had come from Ogaki Kyoritsu Bank — designed SPI (Seria Purchase Index), a model that estimates store-by-store demand for an individual item using conditional probability. The originality was in the translation: the credit-risk assessment a bank applies to a borrower, turned on a single SKU in a single store. When usage reached 100% of stores in 2009, same-store sales turned up, and the argument was settled by the numbers. Ordering stopped being a talent and became a procedure, which meant a new store no longer depended on finding a gifted manager to run it.

The other half of the answer was why anyone would come in. The Color the days format, launched in Yachiyo, Chiba, in November 2007, put interior and kitchen goods into a sparse, muted, select-shop layout designed to break the association between ¥100 and clutter; it reached 240 stores by March 2012 and brought in a largely female customer base that the category had not held before. Between them the two axes reshaped the P&L: variety goods rose from about 75% of sales to 92.6% by the year to March 2012, gross margin to 41.7%, with roughly 500 of some 20,000 items rotated out every month. In the year to March 2009 Seria passed Can Do on revenue to become the industry’s number two.

In June 2014 the founder stepped down and Kawai Eiji became president. Kawai Hiromitsu did not take the chairmanship or any other post; he left. For a mid-sized founder-led company that is unusual, and it made explicit what the systems had already been doing — moving the firm off personal judgment and onto procedure. The new president pushed further in the same direction, adding a multiple-regression model for staffing in 2013 and opening POS data to suppliers, so that when a line softened the maker came back unprompted with a colour change or new packaging. He also noted a harder truth about store openings: as landlords grew selective, expansion had turned from claiming ground into musical chairs, with only profitable chains invited. And the full logic of the ordering system is understood by exactly one person — Kawai Eiji himself. The escape from dependence on individuals stops at the top.

For a decade the yen was a tailwind. A strong currency let Seria upgrade its imported goods, and customers photographed what they bought and posted it, which brought more customers — the president credited both in 2017. After 2022 the same mechanism ran in reverse. Seria sources most of its goods abroad and cannot pass a cost increase through, because the price is the brand; operating margin fell from about 10.6% in the year to March 2021 to about 7.3% two years later. A format built on a constraint discovered what happens when the constraint itself is attacked.

The response was to stop growing outward and start rebuilding inward. New-store expansion was wound back and the capital redirected into the network already in place, with investing cash outflow roughly doubling to about $86.8M (¥12bn), much of it on renewing POS and enlarging distribution centres. Rivals took the other road — Daiso adding ¥300 and ¥500 tiers, Watts launching a separate format — leaving Seria holding a line the industry has begun to question, and betting that a shelf refreshed fast enough can absorb what the price cannot.