The promotion of Satya Nadella to CEO marks a turning point in the history of Microsoft Corp. It's impossible to guess what course Nadella will steer. But by looking back over Microsoft's near 40-year history, we can gain some important insights into what's worked for the company, and what hasn't.
"Everything is becoming digital, and software-driven," says Nadella in a video interview posted by Microsoft. That's an encouraging statement, which acknowledges the company's almost-forgotten raison d'etre: software.
Microsoft was founded on excellence in software, and that excellence has been the biggest factor in its overwhelming success. The robustness of Windows 8 (whatever its design flaws) seems to indicate that Microsoft's programming capability has not atrophied.
The second thing Microsoft was once able to do better than any other software company was to listen to its users. While others pursued grandiose visions (typically, self-serving ones), Microsoft was busy delivering exactly what its customers wanted. That attentiveness has been lacking of late, as indicated by the growing number of unanswered pleas on Microsoft's user forums.
And, finally, what Microsoft has done better, over the years, than any other microcomputing company is to affordably harness the most advanced technologies, and turn them into meaningful capabilities. Other companies had products that were more original, or fancier or cheaper. But Microsoft was best at unlocking raw computing horsepower to solve immediate business needs.
That knack, too, has faded in recent years. Perhaps a look back can show where it went, and help Nadella recapture it.
In the Beginning
A key misconception, popular of late even on the Redmond ‘campus,' is that Microsoft is (or can ever be) a consumer-focused company.
In fact, Microsoft started as a producer of highly technical products aimed at the business and developer communities. Its global dominance was built by serving those constituencies. Nadella's roots in the Enterprise group are therefore somewhat encouraging.
The first product from Bill Gates' and Paul Allen's fledgling company (back in 1975, almost 20 years before Nadella's arrival) was a version of the BASIC programming language for the first true microcomputer, the MITS Altair 8800. This set the tone: not innovation, but robust and timely implementation of advanced technologies, for advanced users.
In its first decades, Microsoft was all about embracing existing technologies, and existing standards. Its first OS was Xenix, an implementation of the cross-platform Unix OS. Its first hardware product was the SoftCard, which allowed the Apple II to run the standard CP/M operating system.
When released in 1981, MS-DOS was largely a clone of CP/M, the dominant OS in business. For a while, it seemed that CP/M might win out. But MS-DOS offered similar advantages, plus, typical of Microsoft, a lot more momentum. Key business applications such as WordStar and dBase swiftly made their way from CP/M to MS-DOS.
Windows of Opportunity
As computing power increased, the shift from 25x80 text displays to dot-addressable graphical user interfaces was inevitable. Like Apple, Microsoft was quick to exploit the pioneering work of the Xerox Palo Alto Research Center (PARC).
But where Steve Jobs aimed for products of visionary design, first the Lisa, then the Macintosh, Microsoft did less-glamorous work under the hood. Windows was a full multitasking system from day one. Icons on the Mac represented files and folders. On Windows 1.0, they represented running applications (presaging today's Taskbar, and Apple's Dock).
Few personal-computer users in 1985 felt a burning need to run multiple programs at one time. But Gates saw the deep elegance of the approach, already universal in the mainframe and minicomputer world.
Gates and hardware partner IBM also saw the huge benefits of backward compatibility. In moving to the Mac, Apple simply left behind its huge Apple II ecosystem. But Windows worked ‘on top' of MS-DOS. It minimized development costs down by leveraging that earlier code, and maximized value by allowing users to run applications for either system.
Windows 3.0 brought the ability to smoothly multitask MS-DOS applications. A few years later, Windows NT shed the MS-DOS underpinnings, but continued to offer a ‘DOS box' that could run the older software. Windows 8 retains this priceless ability, to flawlessly run 30 years-worth of character-based MS-DOS applications.
Microsoft's pragmatic approach was a hit with businesses. The Mac offered elegance that businesses were reluctant to pay for. MS-DOS was a workhorse: the most economical way to run staple business applications such as WordPerfect and Lotus 1-2-3.
Software publishers and business users were skeptical of Windows. Microsoft sold its new OS the hard way: by demonstrating ever-increasing benefits, while keeping the costs low and the upgrade path smooth.
When Lotus and WordPerfect hesitated to deliver Windows versions of their applications, Microsoft stepped in with Excel and Word. Competing in these niches was generally considered futile. Yet Microsoft quickly ended up dominating them to the point of near-monopoly.
Some say Microsoft did this by ‘cheating' in various ways. Not so. Microsoft simply built better software: more functional, more reliable, more cost-effective. Microsoft never went for the quick payout. It took the time to build solid technical foundations, then steadily improved the product.
The whole story is brilliantly (and hilariously) recounted in Merrill Chapman's book "In Search of Stupidity."
In 1993, Microsoft leaped ahead of all competitors with the launch of Windows NT. It was almost a mainframe OS, offering robust multitasking, multithreading, multi-processor support and cross-platform portability. It took Apple the best part of decade to catch up (with OS/X).
Unfortunately, having left the competition behind, Microsoft seemed to lose much of its original impetus. In the early 2000s, it consolidated the gains of Windows NT by migrating its mass-market OS to the same codebase, in Windows XP. Since then, improvements have been incremental at best, negative at worst (e.g. Windows Vista, and, arguably, Windows 8).
Even more obviously, Office peaked with the 2003 version. Subsequent releases, with essentially zero competition to drive them, have added inconsequential features, while demolishing the once-familiar user interface and needlessly increasing complexity.
If only Nadella could re-learn from LibreOffice the value of fundamental simplicity and steady, meaningful evolution.
The Software Game
Consumers gravitated to Microsoft largely of their own accord. A few people started taking their Macs to work, while hordes were bringing MS-DOS (and later Windows) home, to run on their cheap PC ‘clones.'
This growing base of MS-DOS and Windows PCs made an increasingly attractive market for games. Again, Microsoft's pragmatic approach showed its strength, creating a technically-superior gaming platform.
For example, id Software's seminal 3D game, Doom, achieved smooth animation by exploiting the PC's low-resolution 320x200 VGA graphics. The Mac couldn't keep up because, ironically, it was locked to a single, much higher resolution. Its CPU had too many pixels to push.
In the mid-1990s, the flood of MS-DOS and Windows games became so massive, it led to a period of chaos. Simply getting a game to run was a task for computer wizards. Microsoft's solution was it's crowning software achievement: DirectX.
Instead of being designed to run on specific hardware, game software could access the computer through the standard DirectX interface. Meanwhile, instead of worrying about compatibility with every game, hardware (such as graphics cards) needed only a small software driver, which plugged in to the ‘other side' of the DirectX layer.
Microsoft made that layer as ‘thin' as possible, so as not to waste the system's horsepower. No other company could have achieved such a fine balance between compatibility and performance, on fantastically diverse PC hardware. Today, Windows games simply install and run, on all the possible permutations of processors, memory, disk drives, graphics cards and controllers.
Unfortunately, development on DirectX has been all but abandoned, as Microsoft, astoundingly, focuses its efforts on the Xbox.
DirectX created the single largest market for electronic games: about a billion users, compared to the paltry tens of millions offered by even the most successful game console. It's baffling that Microsoft would choose to go head to head on hardware against a fearsome competitor like Sony, when it already owns the most successful gaming software platform imaginable.
Aside from the Xbox, Microsoft's forays into hardware have been rare and narrowly-focused. Mostly, it's been content to dabble in peripherals, especially mice and keyboards, which are obviously complementary to its Windows and Office software.
However, Microsoft has worked extensively with hardware OEMs to create new computing platforms. Its most notable success, dating back to the early 2000s, was the Pocket PC initiative. Again, it was a case of building on previous innovation. But the PocketPC went far beyond the popular PalmPilot ‘personal digital assistant.'
Pocket PCs were manufactured by many companies, including Hewlett-Packard, Dell and others. They ran Microsoft's Windows CE ("compact edition") operating system, which later morphed into the Windows Mobile platform. (Not to be confused with Windows Phone.)
WinCE was incredibly powerful. It had true multitasking, and a true user-accessible file system. It had a user interface that at least vaguely followed Windows traditions. (And could come closer with the aid of third-party software.)
Even now, it's hard to find a handheld device that can offer the same capabilities. Android comes closest, but Windows CE required no tricky jailbreaking to give users total control over their own devices. It also had (still has) a massive legacy of applications.
Yet, as with gaming, Microsoft unaccountably backed away from its robust, successful handheld software platform. While Windows Mobile was going to seed, Apple cleaned up with far less-functional iPods and iPhones. Instead of competing on its own strengths, Microsoft opted to battle Apple on its home turf. (And seemingly ignored Google, which was offering the real successor to Windows CE.)
Microsoft's Zune music player was an early failure: more restrictive than the iPod, with awkward Windows connectivity and an unfamiliar interface. Despite the name, the more recent Windows Phone looks even less like Windows, and offers no compatibility with Windows CE, Windows Mobile or any other facet of the Windows ecosystem.
Can Nadella see the problem? In moving away from the Windows ecosystem, Microsoft loses any strategic advantage. It might as well open a chain of Windows Burger franchises to compete with McDonald's.
The tablet, yet again, was largely a Microsoft creation. (Apple was remarkably resistant to the idea. As recently as 2008, Axiotron was selling premium-priced ‘Modbooks,' created by ripping the screens off MacBooks and replacing them with Wacom tablets.)
Microsoft's Tablet PC program, instituted in 2000, drew support from numerous hardware OEMs, and spawned a range of elegant Windows portables with fold-back screens. The initiative fizzled for two reasons, both indicative of the middle-age decline that was overtaking Microsoft.
Most obviously, Microsoft failed to deliver a properly touch-enabled version of Windows. It added touch sensitivity, but no features or applications to exploit it. Secondly, Microsoft failed to realize that weight was not just incidental, but in fact the dominant criterion for mobile use.
Windows 8 shows that Microsoft has learned little (or drawn entirely the wrong conclusions) from its earlier mobile experience. Windows ‘Metro' offers an interface fully optimized for touch, but, yet again, lacking any continuity with the Windows user interface or the Windows ecosystem.
Microsoft has also not only ventured into hardware, but produced the heaviest, and most expensive, tablets into a market that clearly values light weight and low cost above all other factors. (The Surface Pro 2 is 0.53 inches thick and 910g compared to 0.29 inches and 469g for the industry-leading iPad Air. The Surface 2 starts at $999 while the iPad Air starts at $519 and tops out at $949 fully loaded.)
To be sure, the Surface 2 is a full blown Windows computer. But Nadella needs to ask himself: if that's what consumers really wanted, would they have been buying the iPad and Samsung Galaxy in such vast quantities?
Easing the Transition
What should be apparent is that under Steve Ballmer, Microsoft has forgotten its heritage, and abandoned its core strengths.
In the 1990s, Microsoft launched Windows NT as a parallel upgrade path, instead of forcing everyone to adopt it before the ecosystem was ready. But with Windows 8 Microsoft simply tried to bludgeon desktop users into accepting the bizarrely non-standard Metro user interface.
More importantly, Microsoft countered the success of the iPhone not by revitalizing its superior Windows CE technology, but by fielding devices that are even more restrictive, and (astoundingly!) much less like Windows.
Will Nadella realize the futility of competing with Apple, in hardware markets that Apple knows better than any other company on Earth? Will he realize that, even now, Microsoft could make serious inroads into the much larger Android segment, by leveraging its Windows legacy rather than throwing it overboard?
Does Nadella have the courage to abandon a glowing vision in response to real-world market demand, as Bill Gates did in abandoning his Cairo project to focus on the Internet?
In short, is Satya Nadella the man to restore Microsoft's leadership in software? Or will he simply continue his predecessor's futile attempts to remake Microsoft into a trendy "devices and services" company?
As yet, it's likely that not even Nadella knows for sure. So we can only wish him well... and suggest that if he wants guidance, he need only look to his company's long and remarkable history.