How Microsoft Conquered the PC Era: Bill Gates, IBM, the $75,000 DOS Deal, and the Rise of Windows

In 1980, IBM was one of the world's most valuable companies. Microsoft was a small software business founded by two young programmers who had learned their craft on school computers.

IBM was preparing to enter personal computing. It needed an operating system, and Microsoft did not have one. Bill Gates and Paul Allen found a suitable product at a local computer manufacturer and secured its rights for a total of $75,000.

The amount was small compared with the opportunity. The more consequential decision came when Microsoft negotiated its agreement with IBM: Microsoft retained the right to license the operating system to other computer manufacturers.

IBM would introduce millions of customers to a new computing standard. Microsoft would supply the software to companies competing to manufacture compatible machines.

Over the following fifteen years, Microsoft expanded from programming languages into operating systems, spreadsheets, word processing, presentation software and corporate computing. By the time Windows 95 arrived, the company had become the central software supplier to an enormous global industry.

The story, examined by Ben Gilbert and David Rosenthal in the April 22, 2024 episode of Acquired, Microsoft Volume I, involves much more than an extraordinarily favorable contract. It includes Bill Gates's unusual upbringing, Paul Allen's technical abilities, the economics of software, IBM's changing business model, Apple's graphical computers, Steve Ballmer's enterprise strategy and the competitive pressures that helped establish Microsoft as an industry standard.

It also reveals how a company can become indispensable without manufacturing the products through which most customers encounter its technology.

1. Bill Gates and Paul Allen: An Unusual Beginning

Growing up in a family that understood business

Bill Gates was born in Seattle, Washington, in 1955, the second of three children of William Gates Sr. and Mary Maxwell Gates.

His father's family operated a furniture business in Bremerton, across Puget Sound from Seattle. After military service during World War II, Gates Sr. attended the University of Washington, becoming the first member of his family to receive a college education. He completed undergraduate and law degrees in four years and became a prominent corporate attorney.

He eventually co-founded Preston Gates & Ellis, a law firm that later became part of K&L Gates. His influence extended beyond law: he supported entrepreneurs, helped organize early-stage investors, participated in Seattle's Technology Alliance and served on Costco's board.

Mary Gates came from an established banking family. Her relatives had founded National City Bank, and her father held a senior position at First Interstate Bank.

She built an influential career through institutional leadership rather than a conventional full-time corporate executive position. Her responsibilities included the Seattle Junior League, Seattle Symphony, Chamber of Commerce, Children's Hospital, United Way and the University of Washington's Board of Regents. She also served on corporate boards, including First Interstate Bank, KIRO television and Pacific Northwest Bell.

These positions gave the Gates household unusual access to regional business and political leadership.

Gates later recalled that prominent executives, senators and governors frequently visited the family home. Dinner conversations exposed him to questions of business leadership, institutional responsibility and commercial strategy at an age when most children encountered those subjects only in textbooks.

The young Gates was also exceptionally competitive. Friends remembered his determination to win at academics, sports and other activities. He was an outstanding mathematics student, but his defining characteristics extended beyond intelligence.

Paul Allen, who would become his business partner, described him as brilliant, persistent and eager to demonstrate his abilities.

By approximately age thirteen, Gates was already discussing which company he might eventually lead and what industry he wanted to enter.

This background matters because Microsoft's early decisions were not made by technically gifted programmers who discovered business much later. Gates approached computing with an unusually developed interest in building and controlling a company.

Lakeside School and the extraordinary opportunity to learn programming

In 1968, Gates entered seventh grade at Lakeside School, a private preparatory school in Seattle.

That year, the school's mothers' club raised money to purchase access to a computer.

The arrangement was remarkable for its time. Lakeside obtained a teletype terminal connected through telephone lines to a Digital Equipment Corporation PDP-10 computer in downtown Seattle.

Students typed instructions into a machine resembling a typewriter. Their commands traveled to the remote computer, and responses returned as printed text.

There was no familiar computer screen, graphical desktop, mouse or application menu.

Most computers were expensive institutional systems owned by major corporations, governments and universities. Even smaller minicomputers were beyond the financial reach of ordinary households.

Lakeside gave teenagers an opportunity that experienced professionals in many organizations did not have.

Gates became fascinated with the equipment. Although the terminals were located in the high-school building, he spent extensive time learning programming alongside older students.

One of them was Paul Allen.

The two eventually participated in the four-member Lakeside Programmers Group, an informal organization that aimed to turn its members' programming abilities into commercial work.

Their early language was BASIC, designed to make programming comparatively accessible through readable commands and familiar mathematical operations.

BASIC was not merely an educational novelty. It became an important general-purpose programming language for microcomputers, much as Python later made programming accessible to a broad audience.

At the time, learning to use a computer ordinarily meant learning to program it. The established market for packaged applications did not yet exist.

C-Cubed and the first software royalties

A Seattle company called Computer Center Corporation, commonly known as C-Cubed, provided the students with practical experience.

The business rented computing time to customers. It recruited the Lakeside programmers to identify and document errors in its software.

Rather than paying ordinary wages, C-Cubed compensated them with computer time.

That arrangement was valuable because access allowed Gates and Allen to develop increasingly advanced skills. They moved beyond BASIC into FORTRAN, Lisp, assembly language and machine-level programming.

They also encountered Steve Russell, an influential computer scientist who had developed Spacewar! at MIT, one of the pioneering computer games.

Russell's experience offered a connection to an earlier generation of programmers who had experimented with computers before a consumer software industry existed.

The Lakeside group later obtained a more conventional commercial project: developing payroll and billing software for a Portland-based time-sharing business.

With assistance from Gates's father, the students negotiated compensation through royalties connected to the client's software revenue rather than charging solely for their programming hours.

The arrangement reportedly generated at least $10,000.

That was substantial money for teenagers in the early 1970s. More importantly, it introduced a commercial principle that would shape Microsoft's future: software could be developed once and generate revenue through repeated use.

The customers received useful programs. The programmers could retain an economic interest in their work.

Traf-O-Data and the importance of emulation

After Allen graduated from Lakeside, he and Gates developed a business called Traf-O-Data.

The idea was to help municipalities process information collected from traffic-counting equipment. Governments needed such data for road planning and intersection management.

Gates and Allen believed emerging microprocessors could automate analysis that otherwise required expensive computing resources.

Their proposed machine depended on Intel's early processor technology. But they faced a practical problem: they wanted to begin programming before the required hardware was available.

Allen solved it by writing an emulator on a larger computer.

An emulator reproduces another machine's instructions in software. Programmers can develop applications against the simulated system before running them on the actual hardware.

Allen used processor documentation to reproduce the target environment, allowing the partners to work without waiting for physical chips.

Traf-O-Data reportedly generated approximately $20,000 but never developed into a major commercial business.

Its lasting importance was technical and strategic.

The founders learned how microprocessors worked, gained experience building commercial systems and became comfortable creating software for computers that were not yet readily available.

They would apply the same technique when founding Microsoft.

Harvard and the realization that computing would become universal

Gates entered Harvard in 1973, initially considering a career in mathematics.

Although he had been an exceptional student in Washington, Harvard exposed him to other extraordinarily talented mathematicians. He received a B in the notoriously demanding Math 55 course and gradually shifted his attention toward applied mathematics and computing.

He also met Steve Ballmer, another talented mathematics student whose outgoing personality contrasted with Gates's intensely technical approach.

Allen remained closely involved in Gates's plans. He eventually moved to Boston and worked at Honeywell, keeping himself near Gates while they anticipated the next generation of computing technology.

The crucial development was semiconductor progress.

Processors were becoming faster, more capable and less expensive. Intel's 8080, announced in 1974, strengthened their belief that general-purpose computing would eventually become affordable for ordinary businesses and individuals.

Gates and Allen recognized that exponential technological improvement could create entirely new markets rather than merely improve existing computers.

They envisioned computers becoming everyday equipment in workplaces and homes.

Former Microsoft executive Pete Higgins remembered how implausible this sounded:

"computer on every desk was whackadoo stuff people laughed at it"

The vision eventually became associated with Microsoft's ambition to place a computer on every desk and in every home, running Microsoft software.

At the time, many experienced industry participants expected computers to remain specialized machines used by a minority of workers.

Gates and Allen were betting that the declining cost of computing would overturn that assumption.

2. Why the Computer Industry Was Ready for a Software Company

IBM's integrated model

Before Microsoft, IBM dominated commercial computing.

Its System/360 family represented a highly integrated approach. Customers purchased not just computer hardware, but operating systems, programming tools, maintenance, training and professional services.

A major IBM installation required substantial physical infrastructure and technical expertise. The supplier played a central role in nearly every part of the customer's computing environment.

That arrangement was profitable but also made computers expensive and difficult to introduce into smaller organizations.

IBM's dominance attracted antitrust scrutiny.

In January 1969, the US Department of Justice filed a major antitrust case against the company. Later that year, IBM announced a policy of separately pricing substantial categories of software and services.

The precise date matters: IBM announced this unbundling policy on June 23, 1969, rather than in 1968, as the podcast's chronology suggests. The historical decision is documented in the Computer History Museum's archive of IBM's 1969 pricing policy.

IBM did not suddenly stop supplying integrated computing solutions. But separate pricing helped establish software as something customers could purchase independently of the company manufacturing their computers.

The legal case against IBM continued for years and was ultimately abandoned in 1982.

The broader commercial change survived.

DEC demonstrates how smaller computers create new markets

Digital Equipment Corporation, founded by Ken Olsen, challenged IBM from a different direction.

DEC developed minicomputers that offered less power than the most expensive mainframes but could be purchased by organizations unable to justify enormous installations.

Its original PDP-1, priced at approximately $120,000, was still far beyond the reach of consumers. Nevertheless, it made computing practical for new categories of research institutions, universities and corporate departments.

The strategy illustrated a recurring technological pattern: a smaller, cheaper product can create demand among customers who do not need everything the established market leader supplies.

DEC's machines helped make Lakeside's remote computer access possible.

Gates admired Olsen and regarded DEC as evidence that technological improvements could support entirely new computer markets.

But Gates and Allen anticipated another transition.

Minicomputers reduced the size and price of institutional computing. Microprocessors promised to put much of a computer's processing capability onto a single chip.

If that trend continued, millions of people might eventually own computers. Every one of those machines would need software.

IBM and DEC had established the market for powerful computing equipment. Gates and Allen saw an opportunity to supply the software for a much larger market.

3. The Altair 8800 and Microsoft's First Commercial Battle

A struggling electronics business introduces a personal computer

In January 1975, the magazine Popular Electronics featured the Altair 8800 on its cover.

The machine was developed by MITS, a company founded by Ed Roberts in Albuquerque, New Mexico.

MITS had previously operated in hobby electronics, including model rocketry. By the time the Altair appeared, the company faced serious financial pressure.

The Altair offered a different proposition from institutional computers.

It was advertised at approximately $397 as a kit, placing it within reach of enthusiasts willing to assemble and experiment with their own machines.

The price was possible partly because Roberts negotiated a substantial volume discount on Intel's 8080 processors. The episode reports that MITS obtained processors for approximately $75 each, far below their roughly $360 list price.

The magazine announcement reportedly attracted around 4,000 preorders, creating approximately $1.5 million in advance payments.

These customers were purchasing extremely limited machines.

The Altair's front panel consisted of switches and indicator lights. It had no built-in keyboard, display or familiar graphical interface.

Programming it directly required entering machine instructions.

That limitation created an immediate opportunity for Gates and Allen.

Gates and Allen sell BASIC before writing it

After seeing the Altair announcement, Gates and Allen contacted Roberts and offered to develop a BASIC interpreter.

The interpreter did not yet exist.

Their approach was effectively an early test of customer demand: determine whether MITS would pay for the software, then create it.

Roberts agreed to consider a contract if they could demonstrate a functioning program.

The founders worked rapidly.

Allen constructed an Intel 8080 emulator using Harvard's PDP-10. Gates wrote the BASIC interpreter against the simulated processor.

Their development process meant they could build the software without owning the computer on which it would eventually run.

Allen traveled to Albuquerque for the demonstration. During the flight, he wrote the missing bootloader instructions in octal notation so the program could be loaded into the Altair.

The first demonstration attempt failed. The second worked, allowing the machine to execute a simple arithmetic command.

It was a significant moment: software developed entirely on a different computer had successfully run on the new microprocessor.

Microsoft's own 1975 historical timeline confirms the Altair project, MITS relationship and initial shipment of BASIC.

Micro-Soft becomes a business

Microsoft conventionally dates its founding to April 4, 1975.

Allen is credited with proposing the name Micro-Soft, combining microcomputer and software.

The company initially operated through the founders' developing business relationship rather than the corporate structure it would adopt later. Microsoft's historical records identify a formal MITS licensing agreement on July 22, 1975, the earliest known written use of “Micro-soft” in July and a formal Gates–Allen partnership agreement in February 1977.

Allen joined MITS as its software leader, while Gates continued developing BASIC and increasingly committed himself to their business.

Their initial ownership split gave Gates 60% and Allen 40%, subsequently revised to 64% and 36%.

Gates argued that his greater commitment justified a larger share. Allen's continuing interest in hardware and Gates's determination to concentrate on software also created differences in emphasis.

The founders nevertheless agreed on the opportunity: provide programming tools to the emerging microcomputer industry.

The first licensing contract contains a serious weakness

Microsoft's original MITS agreement gave the computer manufacturer exclusive distribution rights to Altair BASIC.

The financial terms included an immediate payment and royalties on legitimate software sales. Historical records distinguish between approximately $30 per copy for the 4K version and $35 for the 8K version, with additional rights to sublicensing proceeds.

Microsoft's total potential compensation was capped at approximately $180,000.

For a company with almost no revenue, the arrangement offered immediate income and a route into the market.

It also created a serious problem.

MITS controlled whether Microsoft's software would be licensed to other manufacturers.

As competing microcomputer companies emerged, MITS had little incentive to help them by supplying a valuable BASIC interpreter.

Microsoft therefore had a product with broader commercial demand but could not freely sell it.

The contract contained a provision requiring MITS to make its best efforts to license and promote the software. That protection eventually became crucial.

A detailed account in HistoryLink's history of Microsoft's move to Bellevue documents the contract's revenue ceiling, licensing obligations and subsequent dispute.

Piracy reveals a flaw in the business model

Unauthorized software copying created another problem.

MITS reportedly sold approximately 1,000 computers per month, but only a few hundred buyers purchased legitimate copies of BASIC.

Customers could reproduce software much more cheaply than hardware. In the early personal-computer community, copying programs was also part of an established culture of sharing knowledge and technical experimentation.

Gates objected.

His 1976 open letter to hobbyists argued that software development required compensation. If customers copied commercial programs without paying, programmers would have less incentive and fewer resources to create new products.

The controversy occurred before American copyright law had fully developed its treatment of computer programs. Congress clarified their protection through amendments in 1980.

Gates also encouraged talented hobbyists to contribute to legitimate commercial software development.

The economic question was larger than copyright enforcement.

Microsoft was effectively asking individual consumers to make an additional purchasing decision after acquiring a computer.

Yet the computer's usefulness depended heavily on software.

Gates increasingly favored a different arrangement: sell software licenses directly to manufacturers and include their cost in the machine's price.

The manufacturer would pay Microsoft. The customer would receive a functioning computer without needing a separate transaction.

This approach would eventually become one of Microsoft's defining business models.

The 1977 Trinity changes the market

In 1977, three commercially important personal computers appeared: the Apple II, Commodore PET and RadioShack TRS-80.

Together they became known as the 1977 Trinity.

Unlike the Altair, these machines were more accessible to customers who wanted to use computers rather than assemble and program hardware.

They arrived with better distribution, more complete configurations and stronger potential for practical applications.

Microsoft wanted to supply BASIC to these manufacturers, along with companies such as NCR, General Electric and Control Data.

Some potential contracts were reportedly worth approximately $100,000 each.

But MITS's exclusive agreement remained an obstacle.

In May 1977, Roberts sold MITS to Pertec for approximately $6.5 million. The dispute over Microsoft's software rights entered arbitration in Albuquerque.

Microsoft's position was that MITS had failed to fulfill its obligation to promote and sublicense BASIC.

The company was particularly vulnerable during this period because it had interested customers but insufficient freedom to sell to them.

Microsoft won the arbitration toward the end of 1977.

The decision released BASIC from MITS's exclusive control and allowed Microsoft to license the software independently.

Microsoft's 1977 company history records the termination of MITS exclusivity in November and revenue of $381,715 that year.

The importance of the decision extended beyond immediate sales. Microsoft could now become a supplier to the entire emerging computer industry rather than remain dependent on one hardware manufacturer.

Gates tries to make BASIC the common programming standard

After the arbitration, Microsoft expanded agreements with Apple, Commodore, RadioShack and other computer companies.

Apple needed a BASIC implementation with floating-point capabilities beyond the original integer-oriented version written by Steve Wozniak. Microsoft reportedly licensed BASIC to Apple for approximately $31,000 over eight years.

Gates often favored comparatively affordable licenses.

The objective was not necessarily to collect the largest possible payment from each manufacturer. It was to make Microsoft's version of BASIC an obvious and economical choice.

As more manufacturers adopted Microsoft BASIC, developers became familiar with its behavior. Programs increasingly depended on common features and expectations.

Other manufacturers then had reasons to adopt compatible implementations.

Gates described the dynamic as:

"success reinforces success in a growing Market"

The mechanism was straightforward.

A widely adopted software standard becomes more useful because other people already use it. The more users and developers it attracts, the more valuable compatibility becomes.

Microsoft could afford this approach because developing another copy of existing software cost very little compared with manufacturing another computer.

It could license the same underlying work repeatedly without rebuilding it from scratch.

That combination of low reproduction costs and growing compatibility requirements would later become even more powerful with DOS and Windows.

Japan, Seattle and the first substantial revenues

Microsoft expanded internationally unusually early.

In 1978, Japanese entrepreneur Kazuhiko “Kay” Nishi established a relationship with Gates and became Microsoft's principal distribution partner in Japan.

Nishi helped introduce Microsoft's programming languages to Japanese computer manufacturers. By 1979, Japan reportedly accounted for approximately half of Microsoft's revenue.

Microsoft's sales increased from roughly $16,000 in 1975 and $22,000 in 1976 to $381,000 in 1977, $1.3 million in 1978 and $2.4 million in 1979.

The company was still small, but its customer base was expanding rapidly.

Microsoft moved from Albuquerque to Bellevue, Washington, on January 1, 1979, as confirmed by its official 1979 timeline.

Gates valued Seattle's recruiting opportunities, especially the University of Washington's computer science community. He also believed a location outside Silicon Valley would improve employee retention and reduce the risk of confidential information spreading through competing technology firms.

Almost the entire Albuquerque workforce relocated.

The decision helped establish a lasting relationship between Microsoft and the Seattle region. As Microsoft expanded, it recruited generations of local technical talent and contributed to the region's development into a major technology center.

By 1980, Microsoft had moved beyond its beginnings as a small supplier of programming languages.

The next opportunity would change the structure of the entire business.

4. IBM's Personal Computer and the $75,000 Operating System

Project Chess: IBM decides to build a PC in one year

By 1980, microprocessors had become powerful enough to support increasingly sophisticated business applications.

Intel's 8086 marked an important shift from 8-bit to 16-bit processing. More capable processors increased the practical amount of information computers could work with and supported larger, more demanding programs.

IBM could see the personal-computer market developing.

The company had largely allowed DEC to establish the minicomputer category. It did not want to miss another important change in computing.

IBM established an independent development effort in Boca Raton, Florida, under the codename Project Chess.

William Lowe helped initiate the project, and Don Estridge subsequently directed its development.

The team faced a demanding deadline: deliver a marketable personal computer in approximately one year.

Rather than design every component internally, the engineers used commercially available hardware and software wherever possible.

IBM's own history of the original IBM PC confirms that the project was deliberately separated from much of the company's conventional bureaucracy.

This approach accelerated development. But it also meant that many components in the finished machine would be available to competitors.

IBM retained significant technical expertise, including the development of its Basic Input/Output System, or BIOS. Nevertheless, the computer was much more open to outside suppliers than IBM's traditional mainframe systems.

Why IBM approached Microsoft

By the summer of 1980, Microsoft was an established supplier of programming languages for microcomputers.

IBM approached Gates's company for BASIC, COBOL, FORTRAN and related software.

The relationship was logical. Microsoft had deliberately made its languages available across different computer platforms, and IBM needed practical software within a short development schedule.

IBM also needed an operating system.

An operating system coordinates the computer's hardware and provides services through which applications interact with the machine.

Microsoft did not have a suitable 16-bit operating system ready.

Gates referred IBM to Digital Research, the company founded by Gary Kildall.

Digital Research's CP/M was already a widely used microcomputer operating system. It had an established reputation and served as a foundation for many early business applications.

Gates and Kildall knew each other, and their businesses had initially been complementary. Microsoft supplied programming languages, while Digital Research supplied operating-system software.

IBM's arrival changed that relationship.

What really happened between IBM and Gary Kildall?

The failure of IBM's discussions with Digital Research became one of the most frequently repeated stories in computer-industry history.

In the popular version, IBM representatives arrived at Digital Research while Kildall was away flying his airplane. His wife, Dorothy, refused to sign the company's confidentiality agreement, and IBM departed.

The episode describes variations of this account but acknowledges that the circumstances remain disputed.

Some versions say Kildall had a legitimate business appointment rather than a recreational absence. Other accounts emphasize disagreements over the nondisclosure agreement or Digital Research's proposed royalties.

There was also a technical complication: Digital Research needed to deliver a version of CP/M suitable for IBM's new 16-bit computer.

Independent research gives greater weight to the commercial dispute than the familiar airplane anecdote.

The Computer History Museum's examination of early operating-system history describes an impasse over financial terms and notes that Kildall believed an agreement was still possible.

The museum also recounts his unsuccessful attempt to reopen negotiations.

The important distinction is that IBM and Digital Research failed to reach a licensing agreement; the precise personal circumstances are less securely established than the commercial outcome.

IBM returned to Microsoft.

Tim Paterson and Seattle Computer Products

Microsoft found a potential operating system at Seattle Computer Products, a nearby company that manufactured microcomputer hardware.

Its engineer, Tim Paterson, had developed software for the company's Intel 8086-based systems.

The operating system became known as QDOS, or Quick and Dirty Operating System, and later 86-DOS.

Paterson had been frustrated by the delay in obtaining a suitable 16-bit version of CP/M. His system provided similar programming interfaces and practical functionality.

Microsoft first secured licensing rights for $25,000.

In July 1981, it paid an additional $50,000 to acquire broader ownership and distribution rights.

The total was $75,000.

Paterson joined Microsoft and worked with its engineers to adapt the system for IBM's hardware and requirements.

The resulting software became PC DOS when supplied by IBM and MS-DOS when licensed by Microsoft to other manufacturers.

This was not simply a case of Gates writing a revolutionary operating system himself.

Microsoft identified an existing technical solution, acquired commercial control and coordinated development to meet IBM's deadline.

The Computer History Museum's publication of early MS-DOS source code provides particularly useful evidence. It documents Paterson's involvement, the original licensing transaction, the subsequent acquisition and Microsoft's technical work.

The museum also records a later dispute with Seattle Computer Products. SCP sued Microsoft, alleging it had concealed IBM's involvement to obtain the software at an artificially low price. The case was settled outside court for nearly $1 million.

That later settlement complicates the often-celebrated $75,000 purchase: the original price is documented, but it was not the end of the financial consequences.

There were also allegations about DOS's similarities to CP/M. Compatibility and resemblance do not, by themselves, establish unlawful copying, and the historical record distinguishes the operating systems' related interfaces from identical source code.

The IBM contract: ownership mattered more than royalties

Microsoft and IBM signed their landmark agreement in November 1980.

According to the episode, Microsoft's immediate contractual compensation totaled approximately $430,000, divided among testing and consultation, DOS, and programming languages and compilers.

The central financial arrangements were striking.

IBM did not owe Microsoft a continuing royalty on every copy of PC DOS it sold.

That limited Microsoft's direct upside from IBM's own computer shipments.

But Microsoft retained ownership and the right to license DOS to other manufacturers.

This was the more consequential provision.

IBM could popularize personal computing without granting itself exclusive control over the operating system.

Microsoft, meanwhile, could benefit from every manufacturer producing a computer capable of running compatible software.

The two companies were operating with different assumptions.

IBM had reasons to expect that its brand, distribution, engineering and business relationships would allow it to dominate personal-computer manufacturing.

Gates had already watched many different microcomputer manufacturers compete successfully. He anticipated that the market might contain numerous hardware companies rather than one permanent winner.

IBM's history of antitrust scrutiny may also have made nonexclusive software licensing more acceptable.

The arrangement separated hardware production from the software standard that machines would use.

The Computer History Museum's research confirms the essential structure: Microsoft supplied IBM with a nonexclusive license while retaining the ability to distribute MS-DOS elsewhere.

IBM launches its computer

IBM announced the IBM 5150 Personal Computer on August 12, 1981.

The basic machine was priced at approximately $1,565 and came with only 16 kilobytes of memory. Adding necessary equipment significantly increased the practical purchase price.

Its launch was nevertheless an important commercial milestone.

IBM offered customers three operating-system choices: PC DOS, Digital Research's CP/M-86 and the UCSD Pascal system.

DOS was substantially cheaper.

The episode cites approximately $60 for DOS, $175 for CP/M and $450 for Pascal. Historical accounts differ about particular listed prices; the Computer History Museum reports $40 for DOS and $240 for CP/M at the initial launch.

The figures should therefore not be treated as one universally established price schedule.

The economically important fact is consistent: DOS was significantly less expensive than its principal alternatives.

IBM's fixed-payment arrangement with Microsoft also encouraged this pricing. Because IBM did not owe a royalty for every DOS copy, it could make the operating system an attractive purchase.

IBM reportedly sold around 13,500 computers during the initial months and approximately 500,000 over the following two years.

As customers adopted IBM PCs, software developers increasingly targeted the platform.

And Microsoft held the rights to supply that software environment to other manufacturers.

5. Compaq Breaks IBM's Hardware Advantage

Recreating compatibility without copying IBM's BIOS

IBM controlled proprietary BIOS software that coordinated basic interactions between its operating system and hardware.

That appeared to offer protection against manufacturers simply duplicating IBM's computer.

Compaq, founded in 1982 by three former Texas Instruments engineers, recognized a way around the obstacle.

It could purchase standard hardware components, license DOS from Microsoft and independently develop BIOS software that behaved like IBM's implementation.

Compaq engineers used a process known as clean-room reverse engineering.

Engineers studying the original system documented its required behavior. Other engineers then wrote independent code that reproduced the necessary functionality without directly copying IBM's proprietary implementation.

The process made it possible to build legally compatible computers.

Compaq introduced its Portable computer in 1983.

The machine could run software intended for the IBM PC, giving customers access to an established application ecosystem without purchasing IBM hardware.

The Computer History Museum's 1983 chronology records Compaq's legal BIOS reverse engineering, Microsoft licensing relationship and approximately $111 million in first-year sales.

Compaq went public in 1983 and ultimately became the world's largest personal-computer manufacturer in 1994.

Its success encouraged other companies to produce IBM-compatible computers.

Hardware manufacturers compete while Microsoft sells the standard

Compaq's success exposed the weakness in IBM's strategy.

If competing manufacturers could purchase similar components and legally reproduce compatibility, IBM could no longer rely exclusively on its own hardware to control the PC market.

Companies could compete through price, performance, design, distribution and customer service.

Microsoft benefited regardless of which manufacturer sold the computer.

Its licensing agreements with compatible manufacturers increasingly included payments linked to the number of machines shipped.

That differed from Microsoft's fixed-fee agreement with IBM.

The arrangement allowed Microsoft's revenue to grow as the wider industry expanded.

Manufacturers also had practical reasons to include a licensed operating system with each machine. A computer supplied with compatible DOS was more useful to customers than hardware requiring them to locate and install essential software independently.

Bundling therefore helped resolve the piracy and purchasing problems Microsoft had encountered during the Altair era.

Developers targeted DOS because there were many compatible computers. Manufacturers licensed DOS because there were many compatible applications.

IBM had helped create the market. Compaq and other companies expanded it. Microsoft earned revenue from the common software environment.

By fiscal 1984, Microsoft reported revenue of approximately $97.5 million.

The larger commercial pattern was becoming clear: hardware manufacturers competed to sell machines, while Microsoft could serve many of them through one widely adopted operating-system family.

6. The Financial Structure Behind Microsoft's Expansion

Steve Ballmer arrives to manage a growing company

Steve Ballmer joined Microsoft in June 1980, around employee number thirty.

Gates recruited his Harvard friend away from graduate business studies at Stanford because Microsoft's growing organization needed more than programming talent.

Ballmer brought management, recruiting, commercial judgment and an ability to coordinate people across an increasingly complicated company.

He was also an exceptional mathematics student, contrary to the later stereotype that he supplied only sales ability while Gates supplied technical intelligence.

Gates treated him almost as another founder.

He offered Ballmer a substantial ownership stake, initially discussed at approximately 8.5%.

The arrangement created tension with Allen, who had reportedly expected a smaller grant. Gates indicated that the additional ownership could come from his own holdings.

Ballmer became instrumental in expanding Microsoft's management capacity, establishing commercial relationships and eventually creating its enterprise sales organization.

David Marquardt and the only conventional venture-capital round

Microsoft's profitability meant that it did not need the same volume of outside investment as many later technology startups.

Nevertheless, venture investor David Marquardt of Technology Venture Investors developed a relationship with the company.

Ballmer initially screened prospective investors. Several prominent firms showed interest, but Marquardt distinguished himself through questions about Microsoft's strategic future rather than simply offering capital.

His relationship with Gates developed through extended discussions, including a meeting at a University of Washington football game.

Marquardt helped with recruiting, company planning and Microsoft's approach to IBM. He was also associated with the recruitment of Charles Simonyi from Xerox PARC.

Around 1981, TVI invested approximately $1 million for 5% of Microsoft, implying a valuation near $20 million.

That was Microsoft's only conventional venture-capital financing round.

The investment helped formalize corporate governance, alongside Microsoft's incorporation in 1981. A board including Gates, Marquardt and Kay Nishi replaced some of the informal structures of the original partnership.

Microsoft was already commercially successful enough to finance much of its development through its own operations.

The episode reports that around 1980 the company was generating approximately $5 million in revenue and $2 million–$3 million in profit, according to Marquardt's recollection.

Those figures illustrate why outside capital was not the decisive input into Microsoft's early growth.

Paul Allen's illness changes his involvement

In the early 1980s, Allen was diagnosed with Hodgkin's disease.

His treatment required time away from Microsoft, and he eventually stopped working full-time at the company.

His ownership position and relationship with Microsoft continued, including involvement with its board at different points.

The change altered the founding partnership.

Gates remained responsible for much of Microsoft's central technical and strategic direction, while Ballmer assumed growing organizational and commercial responsibilities.

Allen retained a substantial ownership stake in the business he had helped create.

Microsoft becomes a public company

Microsoft moved to its Redmond campus on February 26, 1986.

On March 13, it went public.

The company had approximately $197.5 million in fiscal-year revenue and more than 1,100 employees, according to Microsoft's official 1986 history.

The IPO was priced at $21 per share. Microsoft reports that it raised approximately $61 million.

That is higher than the approximately $45 million cited in the episode's discussion; Microsoft's published historical record supports the $61 million figure for the offering.

Microsoft's valuation was approximately $750 million at the time.

Its ownership remained unusually concentrated for a company of its size.

The episode's ownership estimates put Gates near 49%, Allen around 28%, Ballmer around 7.5% and the venture investor around 6.1%. Microsoft's own IPO retrospective reports Gates retaining approximately 45% following the offering. The estimates should not be treated as identical measurements of the same capitalization date.

Microsoft's IPO was not primarily an emergency fundraising exercise.

The company was profitable, and increasing employee share ownership created securities-law considerations that made a public listing more practical.

The stock offering formalized access to public markets and created liquidity for shareholders.

Microsoft's subsequent appreciation made Gates and Allen enormously wealthy and ultimately created thousands of employee millionaires.

The broader financial lesson is visible in the ownership structure itself.

Because Microsoft could develop software without enormous factories or inventory investments, its founders retained a much larger share of their company than would typically have been possible in a heavily capital-intensive business.

7. Xerox PARC, Apple and the Development of Microsoft Office

The applications that made personal computers useful

The first generation of personal computers appealed largely to enthusiasts who wanted to experiment with programming and electronics.

That began to change with packaged applications.

VisiCalc, introduced in 1979, demonstrated that a personal computer could perform valuable financial calculations through a spreadsheet.

WordStar showed the commercial potential of word processing.

These applications gave business customers practical reasons to purchase computers without learning programming languages.

VisiCalc became especially important to the Apple II. Business users could justify buying the machine because it could perform calculations that had previously required considerable manual work.

Microsoft recognized that applications could become another substantial software business.

It established a Consumer Products Division and began competing in spreadsheets, word processing and other categories.

The division's name reflected an industry still learning how to categorize personal-computer software. Programs that would later become essential corporate tools were initially treated as products for individual computer owners.

Charles Simonyi brings Xerox PARC's experience to Microsoft

Microsoft recruited Charles Simonyi from Xerox's Palo Alto Research Center, better known as Xerox PARC.

PARC had pioneered technologies that would profoundly influence personal computing, including graphical user interfaces, mouse-driven interaction, Ethernet networking and laser printing.

Its researchers included Alan Kay, Bob Metcalfe, Larry Tesler, John Warnock and other influential computer scientists.

PARC's Alto computer, introduced in 1973, demonstrated a graphical desktop years before the Macintosh or Windows became widely available.

The Alto could display windows and graphical information and respond to mouse input, but its expensive hardware made it impractical as a mass-market consumer computer.

Xerox had demonstrated what computing could look like before semiconductor economics made that experience affordable.

Apple became strongly associated with adapting these ideas for personal computers, particularly through the Lisa and Macintosh.

But Apple was not the only company to benefit.

Simonyi brought experience with advanced software design directly into Microsoft, where he contributed to products including Word and Multiplan.

His recruitment demonstrated the value of combining research ideas with commercially available computer hardware.

Why Microsoft lost the first spreadsheet battle

Microsoft's initial spreadsheet product, Multiplan, was intended to run on many different types of computers.

The company believed broad compatibility would create a larger potential market.

Lotus Development Corporation took a different approach.

Its Lotus 1-2-3 spreadsheet concentrated on the IBM PC and took advantage of the newer computer's capabilities.

Lotus offered superior speed, larger spreadsheets and features that made it especially useful for business customers.

The product became a major commercial success.

For a period during the 1980s, Lotus was one of Microsoft's strongest competitors, at times exceeding Microsoft in revenue and market valuation.

Multiplan demonstrated a limitation in Microsoft's strategy.

Supporting many existing computers could create distribution advantages, but it could also prevent software from fully exploiting newer and more capable hardware.

Microsoft needed more than compatibility to defeat Lotus.

It needed to use a technological change to offer something substantially different.

Excel begins on the Macintosh

Microsoft recognized that a graphical spreadsheet could offer advantages beyond the character-based interfaces used by DOS spreadsheets.

Graphical cells, charts, formatting and visual organization could make complex information easier to manipulate.

The company developed Excel around that opportunity.

Remarkably, Excel first appeared on Apple's Macintosh in 1985, two years before its Windows version.

The decision followed Microsoft's early relationship with Apple.

Steve Jobs had introduced Gates to the Macintosh project in 1981. The companies subsequently agreed to cooperate on software development.

Microsoft supplied important applications to the Macintosh, including Word, Multiplan and other business programs.

The original Macintosh launched in January 1984 with only 128 kilobytes of memory. Its graphical interface was impressive, but the hardware imposed serious constraints on application developers.

The later 512 KB model improved matters.

Excel helped demonstrate the practical potential of graphical computing. Instead of merely making traditional spreadsheet commands look more attractive, it could change how users navigated and presented information.

Apple reportedly supported Excel's marketing because compelling applications made Macintosh computers easier to sell.

Gates and Jobs stood together when Excel was announced, a relationship Gates later referenced in Microsoft's 1997 agreement with Apple.

The result was a fascinating competitive arrangement.

Microsoft developed valuable software for Apple's computers while simultaneously building an operating environment that would compete with Apple's graphical platform.

Microsoft Word, PowerPoint and the first productivity bundles

Microsoft Word made its initial DOS debut in October 1983.

Early versions explored mouse-driven interaction before graphical operating systems routinely supplied those capabilities. Microsoft even bundled a mouse with an early Word product.

On the Macintosh, Microsoft began packaging related business applications together.

An early Business Pack combined products such as Word, Multiplan, File and Chart. These arrangements were primarily marketing and pricing bundles rather than deeply integrated application suites.

Microsoft later acquired Forethought, the company behind PowerPoint, in 1987.

PowerPoint added presentations to Microsoft's growing portfolio of business applications.

Word, Excel and PowerPoint eventually became the core products of Microsoft Office.

The first formally branded Office suite appeared for Macintosh in 1989, followed by Office for Windows in 1990.

Microsoft's history of Office confirms the product's development and the importance of Word, Excel and PowerPoint.

The early versions were not as interconnected as their modern successors.

Graphical operating systems, multitasking and common software interfaces made it increasingly practical to transfer information among documents, spreadsheets and presentations.

Office turned separate applications into a recognizable package for workplace productivity.

It also gave Microsoft another advantage.

Customers increasingly depended not only on Windows or DOS but also on applications that stored their work in Microsoft-compatible formats.

8. Windows, Apple and IBM's Attempt to Regain Control

Windows begins as a secondary project

Microsoft announced Windows in November 1983, before the Macintosh reached customers.

The project was originally considered under the name Interface Manager.

Its purpose was to provide a graphical environment for IBM-compatible computers running MS-DOS.

Windows 1.0 did not ship until November 1985.

The first version had significant limitations.

It used tiled rather than overlapping windows, dividing the screen among applications instead of allowing one window to cover another.

It was also not an independent operating system in the modern sense. Microsoft described it as a graphical operating environment running on top of DOS.

Early development had experienced management and scheduling difficulties. Ballmer eventually became involved in coordinating its delivery.

Windows 1.0 did not immediately compete successfully with the Macintosh's graphical experience.

More importantly, Windows was not yet Microsoft's primary operating-system strategy.

For much of the 1980s, Microsoft remained deeply involved with IBM's preferred successor to DOS.

Apple challenges Microsoft's right to use graphical interfaces

Microsoft's graphical software created tension with Apple.

Apple believed that Microsoft's Windows interface used visual ideas and designs derived from Macintosh development.

The companies reached a licensing agreement in November 1985.

But they later disagreed over the agreement's scope.

Apple argued that later Windows versions exceeded the permissions it had granted, particularly as Microsoft introduced more Macintosh-like graphical features.

Microsoft argued that the agreement permitted reuse of specified visual displays in future products.

The dispute became major copyright litigation.

An important independent finding is available in the actual court record.

In Apple Computer, Inc. v. Microsoft Corp., decided by the Ninth Circuit on September 19, 1994, the court upheld Microsoft's position on the central licensing and copyright issues.

The decision was more precise than simply declaring that graphical interfaces could never be copyrighted.

The court distinguished licensed visual displays, protectable artistic expression and general interface ideas that copyright law could not exclusively reserve to one company.

It also recognized the role of Xerox PARC in the development of desktop interface concepts.

The agreement covered much of the visual material Apple challenged, including uses in later Windows products.

The decision limited Apple's ability to prevent Microsoft from developing a broadly similar graphical environment.

Windows could continue expanding without Microsoft having to abandon the graphical interface approach.

IBM develops OS/2 to restore its influence

IBM's position in personal computing had changed significantly since the original PC launched.

Manufacturers such as Compaq could produce compatible machines. Microsoft controlled a widely adopted operating-system family. Intel's processors were becoming increasingly important to the industry's technical direction.

IBM wanted to recover greater control.

One strategy was OS/2, a next-generation operating system intended to replace DOS and support more advanced business applications.

IBM hoped that OS/2 could reinforce its hardware ecosystem and reduce dependence on a market in which competitors could freely imitate the IBM PC standard.

Microsoft agreed to work with IBM on OS/2.

But the partnership contained a conflict.

IBM wanted an operating system closely associated with its own hardware. Microsoft's successful business model depended on supplying software across many competing manufacturers.

Ballmer described the challenge of managing the IBM relationship as:

"riding the bear"

The expression captured Microsoft's dependence on a much larger company whose movements could threaten its independence.

Microsoft publicly supported OS/2 while continuing to invest in Windows.

This was not simply indecision. Microsoft was protecting itself against uncertainty over which operating-system architecture would eventually dominate.

The 286 versus 386 disagreement

The growing gap between IBM and Microsoft became visible in processor technology.

IBM's PC AT used Intel's 80286 processor.

The processor represented a substantial improvement over earlier PCs, but its memory-management and operating-mode limitations complicated advanced multitasking.

Intel's newer 80386 offered more capable 32-bit processing and a stronger foundation for sophisticated graphical software.

Gates strongly criticized the 286's architecture, calling it:

"brain dead"

IBM's product decisions were complicated by the need to protect its broader corporate computing businesses.

A sufficiently powerful and inexpensive PC could perform tasks that customers previously assigned to more expensive systems.

Compaq and other manufacturers were less constrained by those established product lines.

They embraced newer Intel processors and demonstrated that IBM was no longer necessarily the industry's technical leader.

Microsoft increasingly encouraged manufacturers to deliver hardware capable of running more advanced software, regardless of IBM's preferred direction.

OS/2 struggles while Windows improves

OS/2 launched in December 1987 but did not achieve the market dominance IBM had hoped for.

Windows continued developing.

Early versions had limited adoption, but Microsoft's engineers gained experience with graphical interfaces, memory management, device support and application compatibility.

Meanwhile, personal computers became more capable.

The Intel 386 and 486 generations provided substantially stronger performance, making graphical applications more practical.

By May 1990, Microsoft was ready to introduce a significantly improved version of Windows.

9. Windows 3.0 Turns a Backup Plan Into Microsoft's Main Platform

Better hardware finally makes graphical computing practical

Microsoft released Windows 3.0 on May 22, 1990.

Its performance, memory management, graphical interface and compatibility were substantially improved over previous versions.

The system could run familiar DOS applications while introducing users to a more capable graphical working environment.

That compatibility mattered.

Customers had already invested money and time in DOS software. Businesses depended on existing applications, files and procedures.

A new graphical system that required abandoning those investments would have been much harder to adopt.

Windows 3.0 gave users a way to improve their computing experience without immediately replacing everything they already used.

Microsoft's official 1990 chronology identifies advances in memory management and growing third-party application development as important to the release.

The episode reports that Windows 1.0 and 2.0 had reached approximately 5% of the DOS installed base. Windows 3.0 reportedly doubled that penetration within six months.

The company was beginning to convert its enormous DOS distribution into a graphical platform.

Brad Silverberg and the Windows 3.1 team

Brad Silverberg joined Microsoft shortly after Windows 3.0 shipped.

He led development of Windows 3.1 and later became vice president of Microsoft's Personal Systems Division, overseeing Windows 95.

Silverberg recalled that Windows 3.1 initially had a comparatively small team of around 65 people.

At the time, some of Microsoft's more prestigious development efforts involved OS/2 and the enterprise-oriented operating system that would become Windows NT.

But Windows had something those projects lacked: substantial and rapidly increasing customer adoption.

Silverberg also observed a strong difference between IBM and Microsoft's working cultures.

IBM followed formal processes intended to manage large, complex organizations. Microsoft's teams frequently favored direct engineering responsibility, experimentation and rapid decisions.

The differences made cooperation increasingly difficult as Microsoft's engineers became more confident in Windows' future.

Microsoft chooses Windows over IBM

The success of Windows 3.0 forced Microsoft to reconsider its priorities.

Gates, Ballmer and other executives increasingly recognized that Windows could become the company's primary operating-system platform.

The company could no longer justify treating it merely as an alternative to IBM's OS/2.

At a small executive gathering, Microsoft effectively elevated Windows to its central strategy.

The companies separated their operating-system responsibilities around the end of 1990, with IBM taking control of OS/2 and Microsoft concentrating on Windows.

Microsoft's new position was strategically significant.

It no longer needed to follow IBM's preferred hardware and software architecture.

Instead, Microsoft could encourage the entire compatible-PC industry to develop around Windows.

The company also recognized that a graphical operating system needed much more than a user interface. It required reliable interfaces for developers, extensive hardware support and applications compelling enough to persuade customers to upgrade.

A common Windows architecture

Gates outlined Microsoft's broader direction in a 1991 memo.

He described a strategy involving:

"one evolving architecture a couple of implementations and an immense number of great applications"

The distinction between architecture and implementation was important.

Microsoft wanted developers to work with common Windows programming interfaces while allowing different operating-system implementations to serve different customers.

Consumer Windows could emphasize affordability, hardware compatibility and ease of use.

Enterprise Windows could emphasize networking, reliability, security and administration.

The Win32 application programming interface became important to that approach.

Developers could create programs around common software interfaces rather than rewriting everything for fundamentally different computing systems.

Microsoft's own Word and Excel teams also had an important role.

They could demonstrate Windows' capabilities, provide customers with immediate practical value and encourage other developers to invest in the platform.

Windows' future therefore depended on two related products: the operating system itself and the growing collection of applications designed to run on it.

10. Windows 95: When an Operating System Became a Consumer Event

The development of Chicago

During the early 1990s, personal computers were becoming more common in homes as well as offices.

CD-ROM drives, multimedia applications, sound cards, graphical games and improving displays made computers attractive for entertainment and education.

Microsoft needed an operating system that could serve these customers without demanding extensive technical knowledge.

The project that became Windows 95 was developed under the codename Chicago.

Its emphasis was practical.

The product needed to make starting applications easier, improve hardware support, simplify networking and offer a more coherent graphical interface.

It also needed to preserve compatibility with an enormous collection of existing software.

Microsoft had other operating-system projects under development.

Cairo represented more ambitious research into future operating-system capabilities but did not ship as a standalone product.

Windows NT targeted enterprise customers and required comparatively powerful hardware.

Chicago was intended for the mainstream PC market, where price, usability and compatibility mattered immediately.

The development team eventually grew to approximately 360 people.

Silverberg emphasized clear product principles and direct engineering responsibility. Developers were expected to make decisions and solve problems rather than rely on extensive layers of specifications and management approval.

The result was intended to be understandable to people who had never previously used a computer.

The Start menu and a more capable operating system

Windows 95 introduced the Start menu and taskbar as central elements of the desktop experience.

Instead of expecting users to understand command lines or complicated program-management structures, Microsoft created a recognizable starting point.

The product also improved networking, multimedia capabilities, hardware installation and application management.

Its Plug and Play features aimed to make connecting compatible equipment less difficult.

Technically, Windows 95 included substantially more 32-bit operating-system functionality than earlier consumer versions.

It nevertheless retained DOS and 16-bit compatibility components. Describing it as entirely independent of DOS would overstate the architectural break.

That continuity was intentional.

Millions of customers depended on existing DOS programs, applications and device drivers.

Microsoft wanted to improve the computing experience without making those customers abandon their investments.

Windows 95 consequently combined new functionality with compatibility that sometimes made its internal architecture complicated.

The approach helped preserve the enormous ecosystem Microsoft had spent years building.

The launch on August 24, 1995

Microsoft launched Windows 95 on August 24, 1995.

The company organized an elaborate event at its Redmond campus, with Jay Leno joining Gates onstage.

Microsoft licensed the Rolling Stones' “Start Me Up” for the advertising campaign, associating the song with the Start button.

The promotion extended far beyond the United States.

International landmarks were included in launch celebrations, and Windows 95 appeared simultaneously in eight languages.

Microsoft drew on years of investment in software localization to make the release available across major markets.

The campaign was designed for mass audiences.

Only a relatively small proportion of Americans were using the internet in 1995, so television, newspapers and magazines remained central to consumer marketing.

Microsoft presented computers as tools for ordinary people rather than equipment intended exclusively for specialists.

Its spending was enormous. The Computer History Museum estimates the Windows 95 promotional campaign at approximately $300 million.

That figure reflects the scale of the release and Microsoft's effort to turn an operating-system upgrade into a widely recognized consumer product.

Extraordinary sales and growing dependence on Windows

Windows 95 generated remarkable early demand.

The Computer History Museum records more than one million copies sold in the first four days. The episode describes approximately one million copies during the first week and seven million during the first month, although independently published early sales figures vary by measurement period and distribution channel.

Retail sales were only part of the opportunity.

Manufacturers could also supply Windows preinstalled on new PCs, extending Microsoft's reach through hardware sales.

The company's growing dominance reflected both channels.

Microsoft had spent years creating a common operating-system standard. Windows 95 made that standard more useful and more recognizable to individual consumers.

Its commercial performance was substantial.

Microsoft's 1996 annual report recorded revenue of $5.94 billion in fiscal 1995 and $8.67 billion in fiscal 1996.

The company reported net income of $1.45 billion and $2.20 billion, respectively.

For fiscal 1997, Microsoft's financial statements reported approximately $11.36 billion in revenue and $3.45 billion in net income.

These figures provide a firmer measurement than the episode's rounded references to approximately $12 billion in 1997.

The release strengthened Windows' position for years to come.

Microsoft had successfully moved from selling programming tools for early microcomputers to supplying the central operating environment used by millions of households and businesses.

11. Steve Ballmer Builds Microsoft's Enterprise Business

The difference between selling Windows and selling to corporations

Microsoft's early commercial strategy relied heavily on computer manufacturers.

Companies bought DOS licenses, installed the operating system on their machines and sold those machines to customers.

Applications had a different distribution model. Word and Excel could be purchased through software retailers, distributors and other channels.

But major corporations required a third approach.

They needed purchasing agreements, technical support, integration, network administration and confidence that their computer systems would remain reliable.

Microsoft initially had limited experience selling directly to large organizations.

Steve Ballmer played a central role in building that capability.

Former Microsoft executives interviewed for the episode repeatedly credited Ballmer with developing the company's enterprise sales operation.

Office gives Microsoft an unexpected advantage

Microsoft's opportunity inside corporations began with individual employees.

Workers bought or obtained personal computers and used spreadsheets and word processors to perform their jobs.

These computers often operated separately from centralized corporate systems.

Employees created budgets, reports and documents locally, exchanging files through disks, printouts and other methods.

Corporate IT departments had not necessarily selected Microsoft as a strategic supplier.

Yet many employees already depended on Microsoft applications.

This gave Ballmer a powerful argument.

Rather than asking a corporation to adopt unfamiliar technology, Microsoft could point to existing demand inside the organization.

Word and Excel had become valuable to individual workers. Management increasingly needed to support the products employees were already using.

The pattern anticipated what later became known as product-led growth: customers begin using useful software before an organization formalizes the commercial relationship.

Microsoft turned that existing adoption into a foundation for enterprise sales.

Building a direct sales organization

Ballmer expanded Microsoft's relationships with corporate executives, consulting firms, resellers, distributors and independent software vendors.

The company developed dedicated support functions and processes for demonstrating products to business decision-makers.

Microsoft also brought corporate visitors to its campus, where executives and technical staff could explain products and discuss future requirements.

The strategy required new organizational capabilities.

Microsoft needed people who understood complex customer relationships, procurement, training, deployment and long-term support.

It also needed better enterprise software.

Business customers expected centralized administration, security, networking and reliable operation across large numbers of computers.

Microsoft could not satisfy those requirements through consumer-oriented Windows applications alone.

In 1988, the company recruited Mike Maples, IBM's former director of software strategy, to lead its applications business.

That decision reinforced Microsoft's growing ability to recruit experienced executives and operate independently of IBM.

Dave Cutler and the development of Windows NT

Microsoft recruited Dave Cutler from Digital Equipment Corporation in October 1988.

Cutler had extensive experience creating operating systems for demanding institutional computing environments.

His arrival brought expertise that Microsoft's younger PC-focused engineering organization had not developed to the same extent.

Cutler led the development of Windows NT.

Unlike consumer Windows, NT was designed around requirements associated with serious business systems: reliability, security, networking, administration and more sophisticated operating-system architecture.

It was intended to support servers and demanding workstations rather than immediately replace every consumer PC's operating environment.

Windows NT reached the market in 1993.

The project required substantial long-term investment, and its hardware demands made it unsuitable as Microsoft's only near-term operating-system strategy.

Microsoft therefore continued developing consumer Windows separately.

The two product families nevertheless moved toward common application interfaces.

This approach allowed Microsoft to serve different customers while encouraging developers to target a common Windows environment.

NT subsequently became the foundation for important server products, including Windows Server, and influenced the technological systems supporting Microsoft's later enterprise and cloud businesses.

From desktop applications to corporate infrastructure

The enterprise business ultimately expanded Microsoft's role beyond individual documents and spreadsheets.

Products such as Exchange, SQL Server and Active Directory supported communication, databases and centrally managed business systems.

These capabilities changed Microsoft's relationship with corporate customers.

A company using Microsoft Office might depend on familiar document formats.

A company also using Microsoft servers, directories and databases had deeper operational connections to Microsoft's technology.

Replacing the software could require changing applications, employee procedures, administrators' skills and supporting infrastructure.

Microsoft's enterprise strategy therefore reinforced the switching costs already created by Windows and Office.

It also reduced the company's dependence on consumer computer purchases alone.

The combination of individual adoption and centralized purchasing became one of Microsoft's most important commercial strengths.

12. How Microsoft Reinforced Its Position Across the PC Industry

The difference between Microsoft's model and Apple's

Apple and Microsoft pursued fundamentally different approaches to personal computing.

Apple controlled the design of its computers and operating systems together.

That integration could improve the user experience, reduce hardware compatibility problems and allow Apple to coordinate hardware and software features closely.

Microsoft supplied software to numerous independent manufacturers.

This created more hardware choices and greater price competition.

Companies including Compaq, Dell, HP and Gateway could expand PC distribution independently while providing customers with compatible Microsoft operating systems.

Microsoft did not need to finance the factories, component purchases and inventories required to manufacture every computer.

But its model had disadvantages.

Windows needed to work across many combinations of hardware, firmware and device drivers.

That complexity contributed to compatibility problems and instability that were less common in tightly controlled systems.

Apple's Macintosh remained a smaller part of the personal-computer market during this period, even while offering a distinctive graphical experience.

The comparison was not simply about which company had better software. Each business had chosen a different way to organize the relationship between hardware, operating systems, applications and customers.

Developers and manufacturers reinforce one another

The central advantage of an established operating-system platform is that several groups depend on its adoption.

Customers want access to useful applications.

Developers want to write programs for systems with many customers.

Computer manufacturers want to sell machines that run applications customers already expect.

Each group influences the others.

Microsoft's strategy brought these interests together around a common software platform.

Once DOS and later Windows reached sufficient scale, alternative operating systems faced a difficult challenge.

They needed to persuade developers, manufacturers and users to change direction together.

Even an attractive new operating system could struggle without enough compatible applications.

This explains why software compatibility often mattered more than individual technical improvements.

Windows could remain commercially attractive despite shortcomings because customers needed the applications and files associated with its ecosystem.

Document formats become another source of compatibility

Microsoft's advantages extended beyond operating systems.

As Word and Excel spread through workplaces, customers increasingly needed to exchange documents and spreadsheets with colleagues, clients and business partners.

A file created in one application was most useful when other people could open and edit it reliably.

Microsoft's document formats consequently became important to everyday business communication.

Organizations accumulated large collections of files, templates and procedures built around those formats.

Employees also developed skills and habits specific to Microsoft applications.

The cost of switching software was therefore not limited to purchasing a replacement.

Businesses had to consider whether existing documents would work, whether employees needed retraining and whether other organizations could continue exchanging information with them.

This reinforced Office's commercial position even before widespread internet use made digital collaboration routine.

Microsoft learns from competitors

Microsoft did not invent every important category in which it competed.

VisiCalc established the commercial spreadsheet market. WordStar helped popularize word processing. Lotus 1-2-3 became a dominant DOS spreadsheet.

Apple and Xerox had advanced graphical interface technology before Windows became successful.

Microsoft frequently studied these existing products, preserved useful conventions and attempted to improve upon them.

Some Excel keyboard shortcuts reflected earlier Lotus conventions, making the transition easier for experienced spreadsheet users.

Familiarity could be commercially valuable.

Customers did not necessarily want every new application to require learning a completely different way of working.

However, copying existing capabilities was not enough to guarantee success.

Multiplan failed to displace Lotus.

Excel succeeded when Microsoft exploited graphical computing to provide a substantially different spreadsheet experience.

Similarly, early Windows versions struggled until hardware improvements and better software made Windows 3.0 a more compelling product.

Microsoft's competitive record therefore combined imitation, compatibility and investment in new technical capabilities.

The value of repeated improvement

Microsoft became associated with products that improved substantially after their initial releases.

Windows 1.0 was limited. Windows 2.0 made progress. Windows 3.0 established a mainstream graphical platform.

Windows 95 then extended that platform to a much broader consumer audience.

The development pattern reflected an important difference between software and many physical products.

Microsoft could continue improving existing products, release updates and introduce new versions while preserving substantial compatibility with earlier work.

Customer feedback identified weaknesses and helped guide subsequent development.

The company's success depended not only on inventing software but also on maintaining it, improving it and delivering increasingly complex releases.

Over time, Microsoft developed substantial organizational expertise in testing, compatibility and engineering coordination.

One former executive recalled working within an Office organization of roughly 6,000 people that could plan a major software release years ahead and still meet its expected delivery schedule.

Such organizational capabilities became difficult for smaller competitors to reproduce.

13. Licensing Practices and Microsoft's Antitrust Problems

When distribution agreements discourage competitors

The same licensing system that helped distribute DOS and Windows also attracted regulatory scrutiny.

Microsoft used arrangements known as per-processor licenses with certain computer manufacturers during the late 1980s and early 1990s.

Under these agreements, manufacturers could pay Microsoft according to the number of qualifying computers they shipped rather than only the number carrying Microsoft operating systems.

The arrangements could reduce licensing costs and simplify accounting.

But they also created a problem for competing operating-system suppliers.

A manufacturer installing another operating system might still owe Microsoft a fee for the computer.

The manufacturer would then pay both Microsoft and the competing software supplier.

That increased the effective cost of choosing an alternative.

The episode acknowledges this economic effect while questioning how much the contracts contributed to Microsoft's dominance. By then, DOS and Windows already had large installed bases and substantial compatibility advantages.

That distinction is important.

The market power Microsoft had accumulated through customer adoption was not identical to the additional advantage created by restrictive contracts.

What the US government actually established

The Department of Justice began formal antitrust proceedings against Microsoft's licensing practices in July 1994.

The government challenged per-processor licensing, excessively restrictive agreements and other conduct it considered harmful to competition.

Microsoft agreed to settlement terms in 1994. A federal court entered the final judgment in August 1995.

The Department of Justice's original 1994 announcement specifically describes how the per-processor arrangements could require payments to Microsoft on computers carrying competitors' operating systems.

The settlement restricted several practices, including per-processor licenses and certain long-term or exclusionary contractual terms.

The 1995 final judgment established enforceable restrictions.

This evidence confirms that Microsoft's commercial agreements were not simply theoretical antitrust concerns.

They became the subject of government enforcement.

At the same time, the settlement does not establish that every element of Microsoft's operating-system success resulted from unlawful behavior.

The company had substantial network advantages, a large customer base, established applications and widespread manufacturer distribution.

Regulators were concerned with practices that could further restrict competition in a market where Microsoft already held extraordinary influence.

The later internet-era antitrust battles, especially those involving browsers, fell outside the main historical period covered by the episode.

14. Seven Sources of Microsoft's Competitive Strength

The episode evaluates Microsoft's business using seven categories of competitive advantage.

These categories are useful because Microsoft's success cannot be explained adequately by a single product or agreement.

Counter-positioning and industry changes

Microsoft's software licensing model conflicted with IBM's historically integrated business model.

IBM generated revenue through hardware, software and services. Microsoft could earn money supplying many competing manufacturers.

IBM could not simply adopt every aspect of Microsoft's strategy without creating conflicts within its existing businesses.

The transition from mainframes to personal computers changed what customers needed and which companies could supply it most efficiently.

DEC had previously opened new markets through minicomputers. The PC revolution created a still larger reorganization of the industry.

Microsoft positioned itself where those changes increased demand for software.

Economies of scale

Software can be reproduced with extremely low additional production costs.

A program developed for one computer manufacturer could be adapted and licensed to many others.

A new Windows or Excel improvement could benefit millions of customers without requiring Microsoft to multiply its engineering workforce by the same proportion.

The larger Microsoft's customer base became, the more development spending it could spread across sales.

This allowed substantial ongoing investment while preserving attractive economics.

Network effects

Windows became more valuable as developers created applications for it.

Developers were encouraged to support Windows because so many people used it.

Manufacturers then had stronger reasons to supply compatible computers.

These reinforcing relationships helped preserve the platform's position even when competitors offered technically attractive alternatives.

Switching costs

Customers accumulated applications, documents, expertise and business processes connected to Microsoft software.

Changing platforms could require replacing programs, converting files, retraining employees and reorganizing technical systems.

These costs became particularly important in enterprises that relied on multiple interconnected Microsoft products.

Process advantages

Delivering complicated operating systems required extensive engineering coordination.

Microsoft developed expertise in testing, compatibility, documentation, product releases, developer support and working with independent manufacturers.

These activities were less visible than Windows advertising but essential to producing reliable software for millions of different computers.

Branding

Microsoft became widely recognized among both consumers and businesses.

Windows 95 transformed operating-system marketing into a mass-market event.

Enterprise sales and the widespread use of Office reinforced the perception that Microsoft's products were established business standards.

Brand recognition encouraged adoption, although compatibility and ecosystem economics also played major roles.

Strategically valuable resources

Microsoft controlled software rights, code, technical expertise and relationships that competitors could not simply purchase or reproduce.

Ownership of DOS was the clearest early example.

The company subsequently developed important assets in Windows, Office, developer relationships and enterprise infrastructure.

These assets became more valuable as customers and manufacturers built additional products and processes around them.

15. The Organizational Decisions That Sustained Microsoft's Growth

A company built for a rapidly expanding market

Microsoft benefited from one of the most dramatic periods of growth in computing history.

The episode estimates that annual personal-computer shipments increased from approximately 4,000 units in 1975 to nine million in 1986.

That represents roughly 98% compound annual growth over eleven years.

The figure illustrates the extraordinary market expansion surrounding Microsoft's early years.

The company did not need to create every new customer itself.

Semiconductor advances made computers more affordable. Independent manufacturers expanded production. Software developers created new reasons to own computers.

Microsoft benefited by supplying widely adopted software across that expanding market.

This was a powerful combination of technical timing and strategic positioning.

Strategic flexibility under Bill Gates

Gates remained committed to personal-computer software without assuming that one particular product would permanently dominate.

Microsoft developed programming languages, DOS, Macintosh applications, Windows, OS/2 and eventually Windows NT.

Some projects competed with others.

The company accepted that uncertainty rather than abandoning alternatives too early.

Windows provides the most important example.

Microsoft supported IBM's OS/2 strategy while continuing to improve Windows.

When the market favored Windows, Microsoft had an established product and experienced engineering team ready to expand.

This flexibility required both financial resources and management authority.

Gates's substantial ownership helped him pursue long-term technical investments without depending heavily on outside shareholders.

Microsoft's profitability gave the company additional freedom.

Recruiting, retaining and motivating technical talent

Microsoft's early culture placed substantial responsibility on individual engineers.

Employees worked demanding hours and were often directly involved in decisions about products.

Silverberg described developers taking ownership of problems rather than merely implementing detailed instructions prepared elsewhere.

The company attracted ambitious technical and commercial employees who wanted to participate in shaping personal computing.

Seattle's University of Washington provided an important recruiting base.

Microsoft also valued employee retention and institutional knowledge, which partly explained Gates's preference for Seattle over Silicon Valley.

The company expanded internationally while retaining central responsibility for core engineering work.

Local subsidiaries received substantial authority over marketing, sales and relationships appropriate to their markets.

That combination allowed Microsoft to build a common global product business without requiring identical commercial practices everywhere.

Microsoft Research and the search for the next opportunity

Microsoft established Microsoft Research in 1991, under the direction of Nathan Myhrvold.

The organization was intended to explore future computing technologies rather than concentrate exclusively on existing commercial products.

Earlier generations of personal-computer software had benefited from adapting tasks previously performed on larger computers.

As PCs became more capable, that source of readily identifiable opportunities began to mature.

Microsoft needed to investigate new applications, technical architectures and ways of using computing power.

Research therefore became part of the company's effort to remain relevant as existing markets changed.

16. What Microsoft's Success Meant for IBM and the Wider Industry

Market leadership and revenue leadership are different

Microsoft overtook IBM in market capitalization in January 1993.

The milestone attracted attention because IBM had been the defining company of institutional computing for decades.

It did not mean that IBM had suddenly disappeared or become commercially insignificant.

The episode notes that Microsoft's annual revenue did not surpass IBM's until 2015.

IBM's revenue reportedly continued growing until 2012, long after investors had come to regard Microsoft as the more important computing platform company.

IBM retained extensive enterprise relationships, service operations and long-term customer commitments.

This distinction helps explain how technology industries change.

An established company can lose control of a growing technological standard while continuing to serve profitable customers for many years.

Financial markets may recognize the shift in future opportunities long before the older company's absolute revenue declines.

Microsoft built an ecosystem, not just a product catalog

Microsoft's importance extended beyond its own software sales.

Manufacturers made money selling compatible computers.

Independent developers built successful software businesses around DOS and Windows.

Consultants, resellers, distributors and other service providers earned revenue helping customers purchase, install and use Microsoft's technology.

Companies such as Compaq, Lotus and Intuit could establish substantial businesses within the broader personal-computing market.

The ecosystem generated economic activity far beyond Microsoft's own accounts.

That was also important to Microsoft's position.

If other companies could succeed by supporting its platform, they had reasons to continue developing products and services around it.

The company could capture substantial revenue while encouraging a wider industry to expand the platform's usefulness.

Why the IBM agreement remains central

The agreement between IBM and Microsoft was not especially large in immediate financial terms.

Microsoft was paid a modest amount relative to what the PC industry eventually became.

But it retained rights that allowed its software to become an industry-wide standard.

IBM provided credibility and distribution. Competing manufacturers expanded the market. Software developers made the common platform increasingly useful.

Microsoft held a commercially valuable position connecting those activities.

The episode's historical account repeatedly returns to the importance of this arrangement because it shaped the company's economics for years.

IBM's decision to rely on external suppliers helped it launch a PC rapidly.

That same decision allowed Microsoft and other companies to capture opportunities that IBM might otherwise have controlled.

The result was not determined by the operating system's initial technical sophistication alone.

It depended on who owned the software, who could license it, which machines could run it and how customers and developers responded.

Technology leadership remains vulnerable

Gates did not assume that Microsoft's dominance could last indefinitely.

In a 1998 discussion with Warren Buffett, he argued that technology companies could face greater disruption risk than established consumer businesses such as Coca-Cola or Gillette.

Software standards could be powerful, but technological change might eventually undermine them.

Microsoft had already benefited from several such transitions.

Minicomputers challenged mainframes. Microcomputers challenged minicomputers. Graphical operating systems changed the value of command-line software.

Lotus had defeated Multiplan by exploiting a new hardware platform. Microsoft then challenged Lotus through graphical Excel.

The same pattern could eventually threaten Windows.

Microsoft therefore needed to develop new products and capabilities while protecting its existing businesses.

That tension would become increasingly important as the internet changed computing during the second half of the 1990s.

17. The People and Sources Behind the Historical Account

The Acquired episode drew on published histories and interviews with people directly involved in Microsoft's development.

Those interviewed or discussed included Steve Ballmer, Brad Silverberg, Pete Higgins, Mike Slade, Steven Sinofsky, Terry Myerson and David Marquardt.

Their experiences covered different parts of the company.

Ballmer was central to Microsoft's management and enterprise strategy.

Silverberg provided firsthand information about Windows development.

Higgins worked on Excel and the broader Office business.

Marquardt contributed the perspective of an early investor and board member.

Mike Slade had experience with both Microsoft and Apple-related businesses, including work involving Steve Jobs at NeXT and Apple.

The different perspectives are important because Microsoft's history involves overlapping decisions rather than a single uninterrupted plan.

The company had technical disagreements, competing internal projects, contractual disputes and changing relationships with major partners.

Several details are particularly dependent on personal recollection, including private conversations, informal equity arrangements and internal strategic discussions.

Those accounts help explain decisions, but they do not have the same evidentiary status as signed agreements, audited financial statements or court records.

The episode also mentions the Homebrew Computer Club, whose early microcomputer community included influential participants such as Steve Jobs, Steve Wozniak and Bill Gates. The club connected the enthusiasm for personal computing with the disagreements over commercial software copying.

The LGR YouTube channel was recommended as a resource documenting historical computers and software, including DOS and Windows 3.1 systems. Such preservation matters as original machines become more difficult to maintain and operate.

Other conversations in the episode were separate from Microsoft's history. They included André 3000's decision to release flute-centered music rather than conventional rap, the practical use of Ray-Ban Meta smart glasses for hands-free photographs and phone calls, and Summer Health's pediatric messaging service.

These discussions provided personal context but did not materially change the historical account of Microsoft's business.

18. From the Lakeside Computer Room to Windows 95

Microsoft began with two programmers who had unusually early access to computers and an unusually strong understanding of where computing technology might go.

They learned to write software before personal computers became commercial products.

They tested programs on simulated machines because the hardware they needed was not yet available.

They discovered that software could generate royalties, but also that poor licensing agreements could prevent them from reaching customers.

Their first major contractual dispute taught them the importance of controlling distribution rights.

The IBM agreement placed those lessons at the center of the emerging personal-computer industry.

Microsoft acquired a suitable operating system, adapted it for IBM and retained the right to license it to competitors.

When Compaq and other manufacturers made IBM-compatible computers widely available, Microsoft's software became increasingly valuable.

The company then expanded into applications, using the rise of graphical computing to compete with established spreadsheet and word-processing companies.

It developed Windows while collaborating with IBM on OS/2, eventually selecting Windows as its principal platform when technological improvements and customer demand justified the change.

Ballmer built the commercial organization needed to serve large enterprises, while Microsoft invested in advanced operating systems and business infrastructure.

Windows 95 completed another important stage of that development.

A technology that had once required switches, paper terminals and specialized programming knowledge had become accessible through a graphical interface marketed to ordinary households.

Microsoft's position rested on several connected achievements: technical expertise, ownership of essential software, carefully structured licensing rights, manufacturer distribution, compatible applications, enterprise relationships and continuing product development.

Its history also contains clear limitations and disputes. Microsoft's commercial practices attracted antitrust enforcement; partnerships with IBM and Apple produced conflict; and not every product or strategic prediction succeeded.

The company benefited from exceptional circumstances as well as its own decisions.

The central transformation was nevertheless unmistakable.

In the mainframe era, the company manufacturing the computer often controlled the complete system. In the PC era, Microsoft demonstrated that the company controlling a widely adopted software platform could occupy an equally powerful position without manufacturing the computers themselves.

That change reshaped the economics of computing—and established the foundation for Microsoft's next battles over the internet, enterprise software and the future of the operating system.