Company history

The company was incorporated on 9 July 2012 in Minato-ku, Tokyo, as CFO K.K., with $12,533 (¥1m) of capital. Its founder, Daisuke Sasaki, born 1980, was neither an accountant nor a software engineer: a commerce degree from Hitotsubashi in 2004, then Hakuhodo, CLSA Capital Partners, a stint as executive in charge of accounting at the recommendation-engine firm ALBERT in 2007–08, and four years at Google Japan. Two of those experiences produced the idea. His grandfather had run a beauty salon staffed entirely by male hairdressers at a time when the trade was thought to be women’s work — a small operator making an unfamiliar proposition stick, watched at close range. At ALBERT he learned first-hand how heavy paper-based bookkeeping was, and at Google Japan, selling to small firms, he concluded that technology had not reached the way they were actually run. From the outset freee defined its customer as the roughly 6.3 million sole proprietors and small businesses in Japan.

The cloud accounting software freee went live in March 2013, and its design principle was integration. Issue an invoice, track the receipt, make a payment — do the day’s work and the ledger assembles itself, carrying straight through to management analysis, financial statements and the tax return. Bank accounts and credit cards were linked so that statement lines were pulled in and journal entries matched automatically. Where conventional accounting software was a tool for typing numbers into debit and credit columns, freee treated bookkeeping as a by-product of work already done. It won the Launch Pad at the Infinity Ventures Summit in Sapporo that May against eleven other entrants, and 3,300 businesses registered within two months. Pricing began in July: ¥980 a month for sole proprietors filing blue-form returns, ¥1,980 for companies — about half the going rate for comparable cloud services. What the announcement of a “free period” did not say was that the billing mechanism itself was still being built behind it. That July the company took its product’s name and became freee K.K.

Then it left the category it had entered. Cloud payroll software arrived in October 2014 (later renamed freee HR); in May 2015 freee became the first Japanese software to file labour-insurance returns through the government’s e-Gov API; company-formation, My Number administration and a product for financial institutions followed within the year — three or four new work domains annually. In August 2015 it raised $28.9M (¥4bn), an unusual sum for a Japanese startup of the day, because subscription revenue accrues slowly while development and sales spending runs ahead of it: in the year to June 2016 the parent-only accounts showed sales of $5.2M (¥569m) against an ordinary loss of $19.6M (¥2bn), a deficit near four times revenue. On the numbers it looked like a failed business. In a monthly-billing model, the customers already won stay on the books unless they cancel, and it took several more years for investors to price that.

In October 2016 freee connected to Mizuho Bank by API — the first such link with a Japanese megabank. The relationship with regional lenders had started earlier, with a partnership signed with Hokkoku Bank at the end of 2015; its senior managing director, Shuji Tsuemura, said adoption among firms in the Hokuriku region had since grown to about 4,000 companies, and gave the sheer number of customers struggling with their bookkeeping as his reason for the deal. By the spring of 2017 freee was working with seventeen financial institutions and had raised close to $89.2M (¥10bn) in total. In March 2017 the two firms announced a real-time management signal service, in which the bank reads a borrower’s accounting data continuously rather than waiting for an annual set of statements; Sasaki expected it to help bankers understand their customers’ businesses in depth. That same month the cabinet approved an amendment to the Banking Act that effectively obliged banks to open their APIs. The first place a small company lodged its financial data had moved from the bank counter to the accounting software — and the rules caught up afterwards.

In March 2017 freee also added an enterprise plan capable of supporting the statutory audits of listed companies, lifting the ceiling on customer size from sole traders and startups all the way up. It was not a change of market: the company kept small business as its main field, and the enterprise work simply widened the range of firms one product could carry. Its closest brush with a rival came at the same time. In October 2016 freee sued Money Forward in the Tokyo District Court, claiming its MF Cloud Accounting infringed freee’s patent no. 5503795 on automatic journal entry; the question was whether a correspondence-table method and an algorithm generated by machine learning were the same technology. On 27 July 2017 the court dismissed the claim — the same function could be reached by a different route. In this field, a patent was not going to be the barrier to entry.

So freee opened the other way instead. The freee App Store, launched in January 2019, carried applications built by outside developers on the reasoning that 6.3 million businesses have 6.3 million ways of organising their work, and no single vendor can build for all of them. In June a cash-flow navigator, which forecasts funding needs from accumulated accounting data, came out of Freee Finance Lab, a subsidiary set up the previous October. The year to June 2019 closed with revenue of $41.3M (¥5bn) and an operating loss of $25.7M (¥3bn); by that August more than a million business establishments had adopted the software. On 17 December 2019 freee listed on the Tokyo Stock Exchange’s Mothers market at an offer price of ¥2,000, raising about $110.1M (¥12bn) — with roughly 70% of the offering placed with overseas investors, which was rare for a Japanese venture. Sasaki explained that North America was ahead in SaaS and its investors discerning, and that freee had finally grown to a size they could invest in. Domestic cloud accounting penetration was then around 14%: the company went public into a market that barely existed yet.

The pandemic year added products — project management in April 2020, an ordering and procurement service in December — while the customs it exposed, seals and mandatory attendance at the office, worked against remote operation; freee’s own survey found that the commonest reason staff still had to go in was the demands of counterparties. In the spring of 2021 the company raised about $318.8M (¥35bn). By Sasaki’s account it was the largest financing ever done on Mothers, structured in an unusual form that let US investors take part, and it named M&A among its uses of proceeds — rare enough for a listed Japanese company that clearing that hurdle was itself part of the deal. Almost simultaneously, in April, freee bought Sitevisit, whose NINJA SIGN handled electronic contracts; its founder, Masato Kito, described a product that covered drafting, signing, storage and management of contracts in one place for smaller firms.

On 22 June 2021 freee declared a new vision: an integrated management platform where anyone can run a business freely. It rested on three pillars — a cloud ERP that threads across work domains, an open platform for outside developers, and machinery to make transactions between its own users more efficient. The brand was reworked at the same time and the products renamed, accounting freee becoming freee Accounting. The company was no longer selling a collection of individual programs but a single base on which a business could be run. More than 280,000 businesses were then using it.

Two moves followed from that premise. In January 2022 freee issued freee Card Unlimited, becoming the card issuer itself rather than partnering with one, and underwriting the credit from the accounting data it already held — a design that granted higher limits to growing companies, whose server and advertising bills naturally rise, and cut the lag between spending and its appearance in the ledger from three days to one. Sasaki had been refused even a personal credit card just after founding freee, and turned down for a mortgage five years before; the difficulty small operators have in obtaining credit was something he had lived. The acquisitions then came in a run: Likha-iT in the Philippines (2021), the tax software house Mikatus (2022), invoice processing at sweeep and Why (2023), En Japan’s freelancer-management business pasture (2023), Apollo (2024) and YUI (2025) — seven in three and a half years, all of them in the gaps around accounting and payroll: contracts, tax, billing, outsourced labour. What in-house development could not reach at speed, the $318.8M (¥35bn) bought; most of the targets were later absorbed and their products folded into freee’s own names.

Set the results side by side and the losses grew with the revenue. Against $93.8M (¥10bn) of revenue in the year to June 2021 the operating loss was $21.9M (¥2bn), barely different from before the IPO; then $109.6M (¥14bn) of revenue and a $22.8M (¥3bn) loss, $136.6M (¥19bn) and $56.2M (¥8bn), $167.7M (¥25bn) and $55.4M (¥8bn) — revenue up 1.8 times in three years while the deficit more than tripled. The $88.3M (¥12bn) net loss attributable to owners in FY2022 was mostly a single item: a $70.2M (¥9bn) write-down of all fixed assets, software and goodwill included, which freee explicitly declined to attribute to any business underperforming, taken alongside a switch to expensing development costs in full rather than capitalising them. The turn came in the year to June 2025 — revenue $222.5M (¥33bn), operating profit $4M (¥600m), net profit $9.4M (¥1bn) — the first operating profit in the company’s thirteen years, which Sasaki credited to sharper sales and marketing efficiency and to a scale at which it could invest and still earn. Investors did not simply welcome it: the share price fell after the announcement, and he pointed to the gap between the market’s expectations and the company’s own guidance. Nor has freee promised the profit will hold — with acquisitions continuing as a growth strategy, he told shareholders, net income can go negative again. The plan to turn 600,000 businesses’ accounting data into credit decisions and cross-sectional analysis rests on that user base continuing to grow; should it stop, the integration built over thirteen years becomes, unchanged, the constraint.