Masaru Ibuka Built Sony From a Bombed-Out Tokyo Department Store

A Radio Repair Shop in the Ruins

On 7 October 1946, Masaru Ibuka and Akio Morita founded Tokyo Tsushin Kogyo, the company that became Sony. Ibuka was 38 and had spent the war designing precision instruments. Morita was 25, a physics graduate from a wealthy sake family.

The pairing was unlikely on paper. Ibuka was a technical obsessive who had worked on heat-seeking devices and other wartime instrumentation; Morita was the commercial mind, the son of a brewing dynasty who had been groomed to run the family business and instead chose a startup with no capital and no product. What they shared was a conviction that postwar Japan, flattened and short of everything, could compete on precision rather than scale.

That conviction was not a slogan. It was a response to constraint. Japan in 1946 had lost much of its industrial base, its access to raw materials was precarious, and its reputation abroad was poor. Competing on scale against American manufacturers was impossible; competing on the precision of a small number of well-made devices was not. The founders’ bet was that a country short of everything could still be long on craft.

Their first office was a bombed-out department store in Tokyo’s Nihonbashi district. The roof leaked. They repaired shortwave radios and built a crude voltage stabilizer. The company had about 20 employees and almost no capital.

The repair work mattered more than it looks in retrospect. Shortwave radios were the thread connecting occupied Japan to the outside world, and the Americans who brought them in needed people who could fix them. Repair taught the founders how consumer electronics actually failed, where tolerances slipped, which components ran hot. It also paid wages. The voltage stabilizer, unglamorous as it was, solved a real problem in a country whose power grid was unreliable, and it gave the company its first taste of selling a product of its own design rather than a service.

There is a mechanism here worth naming. Repair is diagnostic work: it exposes a company to the full range of ways a product can break, in the hands of users, under conditions the designer did not anticipate. That knowledge is difficult to acquire any other way, and it is exactly what a manufacturer needs before it can design for reliability. The stabilizer, meanwhile, taught the company that a product solving a local, unglamorous problem could find buyers. Neither activity looked like a strategy. Both were.

The context matters here, because the temptation is to read the repair bench as a charming anecdote rather than as a school. Occupied Japan was a place where the normal channels of technical information had been disrupted. Imported goods were scarce and expensive, spare parts scarcer still, and a broken radio was often not replaceable at all. That scarcity pushed repair work toward improvisation: substituting components, rewinding coils, coaxing performance out of equipment that had been designed for a different climate and a different power supply. An engineer who spends a year doing that learns the difference between a design that is elegant on paper and one that survives contact with a user. The founders were not reading about failure modes in a textbook; they were handling them daily, for customers who noticed when the fix did not hold.

The voltage stabilizer deserves a second look for the same reason. A stabilizer is a humble device, but it sits at the boundary between the grid and everything downstream, and building one forces a company to think about input variation, thermal load and the cost of protection. It is the kind of product that sells to institutions because it prevents larger losses, and it taught the young company that reliability could itself be a product attribute rather than an afterthought. Neither the repair service nor the stabilizer was going to make anyone rich. What they did was keep the lights on while the company learned what it was good at.

The Tape Recorder Problem

Ibuka wanted to build a tape recorder. Japan had almost no magnetic tape, and the few imported machines were expensive. So the company made its own tape from paper and magnetic powder. The first Sony tape recorder, the G-Type, went on sale in 1950.

The improvisation was characteristic. Magnetic tape requires a precise coating of iron oxide particles on a stable backing, and the early Japanese product was made by hand-painting a slurry onto paper stock, which produced a surface uneven enough to distort sound. The engineers had to learn coating chemistry, binder formulation and slitting before they could even begin to improve the electronics. The G-Type was, by the standards of imported machines, crude.

The early machines were heavy and unreliable. Ibuka pushed engineers to improve them. By 1951, the company had a model reliable enough for schools and government offices. That business gave Sony the cash flow to pursue bigger bets.

The school and government market was the crucial wedge. Postwar Japan was rebuilding its education system, and tape recorders were a way to deliver standardized audio instruction to crowded classrooms. Institutional buyers cared about durability and service rather than styling, which forced the company to build a support organization and a manufacturing process that could meet repeat orders. The cash flow from that business was modest but predictable, and predictable cash flow is what lets a company take a swing at something as speculative as a transistor.

The counter-evidence is worth acknowledging: the G-Type was not a good machine by imported standards, and the company’s early tape was worse. A modern reader might see this as a failure. It was not, because the point of the exercise was not to match the imports immediately but to build the capability to do so. The school business rewarded durability over refinement, which meant the company could improve its product incrementally while still selling it. That is a luxury a company selling to audiophiles would not have had.

It is worth dwelling on why the tape problem was so hard, because it explains why solving it mattered beyond the recorder itself. Magnetic tape is a materials problem disguised as an electronics problem. The oxide particles have to be uniform, the binder has to hold them without flaking, the backing has to resist stretching, and the slitting has to produce edges clean enough that the tape tracks without wavering. Each of those requirements sits in a different discipline, and a company that learns them learns something general about manufacturing: that the visible product is the last link in a chain of process decisions, and that the chain is where quality is actually made. The hand-painted slurry was a stopgap, but the effort to control it forced the company to build the habits — measurement, batch discipline, tolerances — that later products would depend on.

The institutional market also shaped the company in ways that are easy to miss. Schools and government offices buy in quantities, on schedules, with expectations about service and spare parts. Serving them meant building something that resembled a real organization: a sales channel, a repair network, a production line that could repeat itself. That is unglamorous infrastructure, and it is precisely the kind of thing a startup founded by two men in a leaking building does not naturally have. The recorder business paid for it, and the transistor program inherited it.

The Transistor Bet

In 1953, Ibuka traveled to the United States and learned that Western Electric had licensed transistor technology. He signed a licensing deal for $25,000. The company then spent years trying to make transistors reliable enough for radios.

The $25,000 was a large sum for a company of that size, and the bet was not obviously wise. Western Electric’s transistors were intended for telephone equipment, where they ran at low frequencies and were judged by different standards than a radio receiver demands. Early junction transistors were noisy, inconsistent from unit to unit, and sensitive to temperature. Sony’s engineers had to develop their own manufacturing methods, including the use of a phosphor-doped base and later the mesa and epitaxial processes, to get yields high enough and performance stable enough for a consumer product.

Sony’s first transistor radio, the TR-55, launched in 1955. It was small enough to fit in a shirt pocket, and it made Sony a brand outside Japan. The company renamed itself Sony in 1958.

The pocket claim was a marketing stretch, but the underlying achievement was real: a radio that ran on transistors rather than vacuum tubes drew far less power, generated far less heat, and could be built small enough to carry. That combination opened a market that did not previously exist, and it did so at a moment when Japanese manufacturing was still widely associated abroad with cheap toys rather than precision goods. The TR-55 and its successors helped change that perception, and the name change in 1958 was a signal that the company intended to compete internationally under its own identity rather than as a contract supplier.

The licensing deal itself is instructive. Western Electric was willing to license because it was a telephone monopoly, not a consumer electronics company; the transistors it had developed were a means to an end, and licensing them to others was a way to recoup development costs without entering new markets. Ibuka saw that a technology developed for one purpose could be repurposed for another. The years of work that followed were the price of that repurposing, and they were paid for by the recorder business.

What the transistor years actually required is worth spelling out, because the popular version of the story compresses them into a single bold decision. A transistor for a telephone exchange runs at low frequency and can tolerate characteristics that would be unacceptable in a radio. A radio receiver needs gain, selectivity and low noise across a band, and it needs those things consistently across thousands of units. Getting there meant not one invention but a series of manufacturing problems solved in sequence: controlling doping, stabilizing the base layer, improving yield, testing at scale. The phosphor-doped base and the later mesa and epitaxial processes were steps in that sequence, each one buying a little more consistency. None of them was glamorous. Together they turned a licensed laboratory curiosity into something that could be stamped out by the thousand and sold to ordinary buyers.

The timing was as important as the technology. Japanese goods abroad carried a reputation for cheapness, and a small, well-made radio was a way to argue against that reputation with a product rather than a slogan. The TR-55 was not the first transistor radio on the market, but it was small enough and reliable enough to change what buyers expected a portable radio to be. The renaming in 1958 consolidated the argument: the company was no longer a contractor making parts under someone else’s label but a brand asking to be judged on its own name.

The Walkman and the Empire

The Sony Walkman, launched in 1979, was not Ibuka’s idea. It came from Morita and engineer Kozo Ohsone. But it was built on the culture Ibuka created: engineers given room to make small, precise consumer products.

The Walkman is often told as a story of a single insight, but the mechanism was more mundane. Sony had a portable tape player, the Pressman, that was used by journalists; Morita wanted a version for music, stripped of the recording function and the speaker, played through headphones. The engineering challenge was not inventing a new format but miniaturizing and ruggedizing existing parts while keeping playback quality acceptable and battery life usable. The company’s accumulated competence in small motors, tape transport and low-power electronics, built up over two decades, is what made the product feasible.

Ibuka retired as chairman in 1976 and died in 1997. By then Sony was a global electronics empire. The company’s later struggles with smartphones and Hollywood are a different story. The original rise was built on repairs, tape, and transistors in a ruined city.

The Walkman also illustrates how a company’s internal culture can outlast the specific problems that created it. The habits formed at the repair bench and in the tape coating room — attention to tolerances, willingness to strip a product down to what the user actually needs, comfort with small mechanisms — were available to the engineers who built the portable player. Morita’s contribution was to frame the product in terms of a use case rather than a specification: music, on the move, through headphones, without the recording function that most buyers never used. That framing only worked because the company could execute it, and the execution rested on two decades of accumulated competence.

What the early history shows is a pattern rather than a prophecy. Each stage of the company’s growth depended on a capability earned in the previous one: repair work taught failure modes, tape manufacturing taught materials discipline, institutional sales taught service and repeatability, and the transistor license converted those habits into a product that could travel. None of it required a grand strategy document. It required founders who were willing to make their own tape because the alternative was to wait.

For the reader, the lesson is less about genius than about sequencing. Ibuka and Morita did not begin with a vision of a global empire; they began with a leaking roof and a repair bench, and they took the next step that their existing skills made possible. The transistor bet looks bold in hindsight, but it was financed by the unglamorous school recorder business. The Walkman looks like a flash of inspiration, but it rested on years of miniaturization. The pattern is worth remembering whenever a small operation seems too marginal to matter: the marginal work is often where the next capability is learned.

It is also worth remembering that the pattern is not a guarantee. Many companies do repair work and never build a transistor radio; many make their own tape and never sell a Walkman. What distinguished this one was not the sequence itself but the willingness to treat each unglamorous step as a chance to learn something that would matter later. The repair bench was not a detour from the real business. It was the real business, for a while, and it taught the company what it needed to know.

There is a further caution in the same history. The capabilities that carried Sony through its first three decades were specific to a world of precision mechanics and analog electronics, and the company’s later difficulties suggest that a capability earned in one era does not automatically transfer to the next. The pattern described here is a way of building competence, not a permanent shield against change. But that is an argument for taking the marginal work seriously, not against it: the alternative to learning on the repair bench is not a grand strategy, it is waiting, and waiting is the one thing the founders of Tokyo Tsushin Kogyo could not afford to do.