Canon — Company History

A domestic luxury camera funded by an obstetrician

In November 1933 Yoshida Goro (吉田五郎), an engineer working on projectors, and Uchida Saburo (内田三郎), who worked in the securities business, set up Seiki Kogaku Kenkyusho — the Precision Optical Instruments Laboratory — in a room of a block of flats at Roppongi in Azabu, Tokyo. The purpose they set out was the research and manufacture of precision high-grade 35mm cameras, a field then held by imported goods. The money came from Mitarai Takeshi (御手洗毅), head of obstetrics and gynaecology at Seibo Hospital, who had attended the birth of Uchida's wife. In 1935 the painstaking research finally reached a marketable product, named at first Kwanon after the thousand-armed goddess of mercy, but soon changed to Canon — a word meaning scripture, norm and standard, chosen to carry the ideal of making a camera that would itself be the standard of excellence. In 1936 a new plant at Meguro took the work into volume production, and in August 1937 the firm was reorganised as Seiki Kogaku Kogyo Co., Ltd. with capital of ¥1 million. In 1939 it began making its own lenses, and in the same year completed the X-Ray Canon, Japan's first indirect X-ray photography unit, stepping into medical equipment as well.

In October 1945, the month after the defeat, the company was quick to restart production. The Canon S that it sent out the following year, 1946, was well received among Occupation officers and men and among buyers visiting Japan, and this was the beginning of the Canon name being recognised abroad. In this period Mitarai Takeshi gave up his post as an obstetrician to devote himself to management, and set as the purpose of the business the task of bringing to fruition a domestic camera able to stand against foreign goods. In September 1947 the trading name was changed to Canon Camera Co., Ltd., and in May 1949 the shares were listed on the Tokyo Stock Exchange, opening a route to funding from the capital market. Of his 1950 tour of Europe and America Mitarai said that it came down to how little Japanese goods were trusted and how low the moral sense of Japanese goods was; I felt an impulse that would not let me rest until I had roused the industry (Shin Nihon Keizai, 20 Oct 1950) — he took the low standing of Made in Japan as the precondition for exporting.

The tenfold bet on the Shimomaruko plant that carried the export business

In 1951 Mitarai Takeshi signed a sole overseas export agency contract with Jardine Matheson, the British trading house, and with a loan of US$500,000 alongside it set out on exporting in earnest. In November of the same year he brought the head office and the factory together at Shimomaruko in Ota-ku, Tokyo, as groundwork for future growth. The sum committed was $555,556 (¥200m), ten times the ¥20 million of capital the company then had, and far above the normal level of capital spending. A service centre was opened in Ginza, Tokyo at the same time, giving domestic customers a point of contact. The finished plant was described at the time as a leap from a wooden backstreet workshop to a modern factory (Diamond, 23 Oct 1956), and this production design — specialising in high-grade cameras and putting in scale ahead of demand so that quality and quantity were secured together — became the physical base on which the export business rested from then on.

Specialising in high-grade cameras kept the ratio of profit to sales above 10 per cent, and the ¥200 million investment was recovered in three years. Meguro Seiki Seisakusho was founded in December 1952 and Chichibu Eikosha in May 1954, putting the supply of parts inside the company's own group. In October 1955 a New York branch was opened to put a sales organisation on the ground, giving direct access to the American market. In September 1957 Canon Europe S.A. was placed in Switzerland as the sole agency for Europe, so that both North America and Europe were held through Canon's own sales network. Under Mitarai Takeshi's rule that volume production should be held back until a thoroughly satisfying product existed, exports were not hurried until quality was certain, and the sequence of moving into mass production and export only once preparation was complete became settled practice.

From 1956 the company brought out in quick succession the Canon 8T 8mm cine camera, the Canon Zoom 8, an 8mm projector and lenses for television cameras, widening the range. In 1961 it sent out the Canonet, a forerunner of the EE camera, which together with the flagship Canon 7 Model F drew the attention of the market. In October 1964 it put an electronic desktop calculator on sale, entering the office-equipment field in earnest. In April 1966 it established Canon U.S.A., Inc. and switched to selling through a local subsidiary. As of 1968 sales were divided roughly a quarter each between high-grade cameras, mid-range cameras, 8mm products and non-camera special equipment, and exports accounted for close to half of all sales. Canon was the company that brought mass production and mass selling into the industry ahead of its rivals, and within the special-equipment division the IC-based electronic desktop calculator was growing.

From the exit from calculators to Kaku Ryuzaburo's reforms

In 1967 Mitarai Takeshi declared that the company must brandish cameras in the right hand and office machines, optical products and special equipment in the left, and on top of that grow exports greatly, setting out a policy that made cameras and office equipment the two wheels of the business. In February 1968 Canon Business Machines Sales Co., Ltd. was established, and in April of the same year the NP system was developed, taking the company into plain-paper copiers. In March 1969 the trading name was changed to Canon Inc., dropping the word camera from the company name. In March 1970 it announced semiconductor production equipment, starting the line that would later lead to industrial equipment. In November 1971 Canon Camera Sales and Canon Business Machines Service were folded into Canon Business Machines Sales, and the domestic selling function was unified as Canon Sales Co., Ltd.

Xerox had commercialised the plain-paper copier in 1960 and had kept others out with a patent wall built over twenty years. Canon too had at first commercialised the Electrofax process, which used coated paper, but because the paper was coated with zinc oxide the ground never came out white, and the paper itself was thick and heavy. In 1962 a product research section was created; of its seven members, three including Tanaka Hiroshi (田中宏) took charge of copiers and worked through the Xerox patents, only to establish that no small evasion would make a product possible. Tanaka then turned to cadmium sulphide, a material he had been studying for the electronic control of cameras. Its resistance was low and it could not hold a charge, so it had been given up as a material for electrophotography; but by integrating an insulating transparent film on its surface to hold the charge, separating the function of charging from that of retention, the defect was turned into an advantage. Combining two materials optically was the home ground of a camera maker, and the process was named the NP system.

An E Project of twenty-three people started in 1965, but practical application was not easy, and the copier division ran up cumulative losses of $2.8M (¥1bn). Strong voices inside the company called for withdrawal, yet management judged the business promising over the long run, moved the division into the Central Research Laboratory and let development continue at head-office expense. The first machine, the NP-1100, reached the market in 1970, in the eighth year from the start of the work. Because Canon had not grown a sales force of its own for copiers, it chose to license its patents freely to foreign firms eager to enter and take royalties in return. From 1977 royalty income on copiers exceeded royalty payments, and from 1980 the payments were nil. Establishing the patents had taken tens of thousands of pages of documents and several hundred million yen in legal fees.

The Hewlett-Packard OEM deal that opened a 70 per cent world share in laser printers

Calculators, meanwhile, had their margins destroyed by the price war of the 1970s, and the market turned a sceptical eye on how loosely the diversification had been narrowed. At the start of 1972 the company's course was described as diversification without a decisive card (Shukan Toyo Keizai, 15 Jan 1972). In June 1975 Canon passed its dividend and withdrew from calculators. Amid this, and with the sudden death of president Maeda Takeo (前田武男), in 1977 the managing director Kaku Ryuzaburo (賀来龍三郎) was raised to the presidency over the heads of the vice-president and the senior managing director. Kaku said openly that the previous management had been lax, and laid down a divisional system that made cameras, office equipment and optical products each answer for their own profit and loss. Immediately after his appointment a severe view was also being expressed — that making office equipment comprehensive was Canon's task, and that the question was whether it could differentiate its products with high added value (Shukan Toyo Keizai, 30 Jul 1977).

The copier business expanded rapidly from the late 1970s, and in 1982 its sales passed $401.5M (¥100bn). Seeing signs that growth was falling to single digits, in the same year the company put on sale the Mini Copier, small and cheap enough to sit on a desk, opening the market for personal copying. In May 1975 it had succeeded in developing a laser printer, and in 1985 it concluded a business co-operation agreement with Hewlett-Packard of the United States and supplied laser beam printers to it on an OEM basis. Rather than building a branded sales channel of its own, it handed the front of the sale to the partner in order to take volume, and the world share of laser beam printers reached 70 per cent in 1990. Sales to HP later grew to a scale of ¥611bn. The bubble-jet recording process developed in October 1981 bore fruit in the BJ-10v, a printer for individual users launched in 1990.

Overseas production spread in parallel: Oita Canon in February 1982, Canon Virginia, Inc. in the United States in November 1985, Canon Opto in Malaysia in December 1988, Canon Dalian Business Machines in China in September 1989 and Canon Hi-Tech in Thailand in August 1990. In the ten years from 1977, when Kaku became president, Canon moved the weight of its business from camera maker to office-equipment maker. In 1990 it was reported to have grown into a blue-chip company ranking fifth in Japan by consolidated recurring profit, with sales above $6.9B (¥1tn) (Nikkei Business, 1 Jan 1990).

Digitisation, and a structure with four-fifths of sales overseas

In September 1995 Mitarai Fujio (御手洗冨士夫), a nephew of the founder, became president and representative director. He had left the faculty of law at Chuo University in 1961 to join Canon Camera, and from 1966 had been seconded to Canon U.S.A., spending more than twenty years in the United States. At the time of his appointment consolidated sales were $22.2B (¥2.09tn) and net profit $584.8M (¥55bn). Mitarai changed the yardstick from the profitability of each division to the profit and cash flow of the whole consolidated group including subsidiaries, and pressed on with withdrawal from businesses whose returns were hard to see and with the concentration of research and development. He held up self-reliance — keeping technology and sales channels in the company's own hands — and moved pay and treatment away from seniority towards ability.

The first thing he took in hand after his appointment was production: within about four years he did away with the belt conveyors at all fifty-four plants and switched to the cell system. The dwell time during which work in progress lay untouched by human hands shrank in the accounts from twenty-three days to five, and parts inventory from about three days to five or six hours. Compressing work in progress and parts inventory cut the working capital of the plants to a third, freeing something of the order of $1.8B (¥200bn). About twenty thousand metres of belt conveyor were scrapped, 720,000 square metres were freed across the fifty-four plants, and the outside warehouses under contract fell from thirty-seven to twelve, cutting $26.4M (¥3bn) a year in costs. The cost ratio, about 62 per cent in 1995, had come down to 50 per cent by 2003. Mitarai described the process as nothing but persuasion, going round to explain to each plant manager the success of the Nagahama plant where it had first been introduced (Shukan Toyo Keizai, 27 Dec 2003).

As digital cameras spread in earnest from the late 1990s and the shift away from film accelerated, Canon ran in the leading group in digital single-lens reflex cameras with the EOS series and held the earnings base of its imaging business. In the 2000s the ratio of overseas sales reached 80 per cent, and a structure was in place for operating globally on the two pillars of office equipment and imaging. In September 2000 the company listed on the New York Stock Exchange (delisted in March 2023). Keeping production centred on Japan, it also widened its Asian bases — Canon (China) in March 1997, Canon Vietnam in April 2001 and Canon (Suzhou) in September of the same year — raising its resistance to swings in currencies and geopolitics.

The bet on acquisitions in medical equipment and network cameras

Consolidated sales grew to $38.1B (¥4.48tn) with operating profit of $6.4B (¥757bn) in the year to December 2007, then fell to $34.3B (¥3.21tn) with operating profit of $2.3B (¥217bn) in the year to December 2009, across the global financial crisis. Office multifunction machines in the advanced economies had worked through their diffusion, and replacement cycles were lengthening. In November 2009 Canon announced that it would make Océ, the large Dutch printing-machine maker, a consolidated subsidiary through a tender offer. At €8.6 a share, some €730 million in all, it was the largest acquisition the company had made to that point. Océ held one of the leading positions in commercial printing in Europe, was strong in document and industrial printing systems and in high-speed wide-format printing, and did not overlap Canon's copiers and wide-format printers. It marked the turn of a company that had held to developing everything itself into one that bought other firms' technology and channels.

Océ became a subsidiary in March 2010, and acquisitions continued after it. In April 2014 Canon took over Molecular Imprints of the United States, gaining nanoimprint technology for semiconductors, and in July of the same year brought in Milestone Systems of Denmark to fill a thin patch in video management software. In April 2015 it acquired Axis of Sweden for about $2.5B (¥300bn), taking the then world lead in network cameras. In December 2016 it bought Toshiba Medical Systems for $6.1B (¥666bn) and entered medical equipment in earnest; that company was the world's fourth-largest maker of diagnostic imaging equipment after GE of the United States, Siemens of Germany and Philips of the Netherlands, and stood first in Japan and third in the world in CT. Inside Canon these four — industrial equipment, commercial printing machines, network cameras and medical equipment — were called the four pillars, and the money spent on acquisitions from 2010 came to more than $9.2B (¥1tn) in all.

Mitarai explained the run of acquisitions by saying that both cameras and office equipment have lost their power to grow. We have to change the portfolio, adding that there was no contest to be won by the leisurely method of developing one's own technology into a new industry (Shukan Toyo Keizai, 22 Apr 2017). Divisional sales of office equipment, which had grown to just short of ¥3tn in 2007, later fell below ¥2tn, and the divisional operating margin dropped from 22.3 per cent in 2006 to less than half that. Shipments of the highly profitable single-lens reflex cameras have roughly halved from their peak in 2012. Against that, the annual operating profit of Axis and of Toshiba Medical each stayed at around $91.9M (¥10bn), a long way from purchase prices of ¥300bn and ¥665.5bn. These were acquisitions that carried large amounts of goodwill, and the structure was one in which impairment would be forced if the expected earnings did not come through.

Shareholders' eyes on a long tenure, and the impairment in medical equipment

Mitarai had been a director since 1981, moving between president, chairman and CEO while remaining at the centre of management. At the annual general meeting of 30 March 2023 the vote in favour of his re-election as a director stopped at 50.59 per cent. Support, about 90 per cent up to 2021, had fallen to 75.28 per cent in 2022 and dropped further that year, leaving a wafer-thin re-election only just above the level needed to carry. All five directors, the outside ones included, were men, and ISS, the proxy adviser, recommended a vote against on the grounds that no woman sat on the board, with foreign institutional investors following. Every reason given against that could be confirmed was the lack of diversity on the board. Trading was firm: in the year to December 2022 sales passed ¥4tn for the first time in five years, at $30.7B (¥4.03tn). When the first woman outside director was brought in the following year, in 2024, the vote in favour returned to 90.86 per cent.

On 30 January 2025 Canon announced that in its results for the year to December 2024 it had booked an impairment loss on goodwill of $1.1B (¥165bn), centred on its medical-equipment subsidiary. Sales set a record at $29.8B (¥4.51tn), yet net profit fell by four-tenths from $1.9B (¥265bn) the year before to $1.1B (¥160bn), the first decline in four years. Since the acquisition the medical-equipment business had grown its sales above ¥500bn, but its operating margin had stalled at 4 to 5 per cent. A weak yen and higher raw-material costs worsened the operating environment for medical institutions in Japan, and the Chinese market on which growth had been counted slowed sharply, so that when the future plans were revised conservatively the value of the business fell below its book value. In February 2024 Canon set up a medical business innovation committee, reversing the policy of leaving the management of acquired companies to each of them, and is bringing research and development and head-office functions together in Ota-ku, Tokyo.

On 29 January 2026 Canon announced that Ogawa Kazuto (小川一登), director and executive vice-president, would be promoted to president and COO. Mitarai gives up the presidency while continuing to direct the group as chairman and CEO. Ogawa takes office after the annual general meeting of 27 March, and it was the third time — after 2006 and 2016 — that Mitarai had faced a press conference on handing over the presidency. Ogawa joined the company in 1981 and came up through overseas sales, spending some thirty years posted abroad in Singapore, Hong Kong, China and Canada among other places, and as president of Canon U.S.A. carried the recovery from the pandemic. Against Mitarai's long tenure — forty-four years as a director and about thirty at the top — voices had been calling for a generational change, including over the ageing of the management core, with the CFO and CTO in their seventies and eighties as well. Consolidated sales for the year to December 2025 were $30.9B (¥4.62tn) and net profit $2.2B (¥332bn), recovering from the previous year in which the impairment was booked.