Schemers
Volume Number: 11
Issue Number: 8
Column Tag: Inside Info
By Chris Espinosa, Apple Computer
All summer the news has been full of big deals and grand schemes in the computer
industry. The most notable ones are the failed Microsoft acquisition of Intuit, and the
IBM purchase of Lotus Development.
(It’s funny that, five years ago, IBM’s purchase of Lotus would have run into
major anti-trust problems, but Microsoft’s bid for Intuit might have gone unnoticed.
But that just shows how much changed from 1990 to 1995).
The interesting thing about these deals is that they were seen as not just normal
bids to broaden a product line or buy market share, but rather as the first play in
intricate schemes. Microsoft’s acquisition of Intuit, if you recall, was not just a trade
of a losing position in home finance software for a winning one. It was seen as the
foundation for a massive Pynchonian scheme to control electronic banking and
commerce, wired into the Microsoft Network, credit cards, video-on-demand, new
forms of software distribution, and even the nature of money itself in the new century.
Pretty heady stuff for a checkbook program.
And IBM’s Lotus move was not simply to buttress a traditionally thin applications
offering from the world’s largest computer company: it was a strategic play to control
corporate client/server application development and internal communications, and
place IBM mainframes, servers, and communications back in their rightful role as the
backbone of the world’s corporate information structure.
Doubtless some of the executives of the two firms were harboring dreams like
these. And maybe these schemes swayed the finance types and board members who
normally cast a cold eye on big-money transactions. But the reality of the industry is
that for every big scheme that works, plenty fail.
The failures are quickly forgotten. Perhaps the biggest was IBM’s System
Application Architecture (SAA), a grand scheme to unify the PC, mini, workstation,
and mainframe lines under a single family of operating systems, user interfaces, and
communications protocols. The downfall of SAA was its largeness. So many
constituencies wanted the standards to address their needs that the standards were
simply generalized to cover everything that existed, and all the benefits of convergence
were lost. Apple, too, had a grand scheme in 1992 called “client/client/server,”
where desktop Macs, PDAs like Newton, and Unix-based servers would all be united
under a common set of protocols and APIs.
Microsoft is the current king of schemes, some fruitful (like the Microsoft
Office) and some less so (like the ACE initiative, Microsoft At Work, Windows for Pen
Computing, OS/2 as the replacement for DOS, Windows NT as the replacement for DOS,
and Windows 95 as the replacement for DOS).
But there have been three big schemes these companies that have worked
tremendously well. And what they have in common is that they were just simple ideas
that were only seen as schemes after the fact.
IBM’s big win was making the IBM PC architecture dominate 80% of the
computer market in about two years, displacing the CP/M and S-100 Bus architecture
and blowing by the Apple II, Commodore PET, and other competitors. They did this by
doing things common to their competitors but entirely out of character for IBM. They
bought their OS from someone else; got others to write applications for them; and even
published the circuit diagrams of the computer. The scheme was to dominate by having
the most hardware and software add-ons. What really happened, of course, was that
Compaq cloned their machine, as did hundreds of others. IBM ended up with less than
ten per cent of the IBM market. But the market was many more times larger than it
would have been if IBM had sold yet another closed box.
Apple’s successful scheme was to avoid IBM’s mistake. With the Macintosh, we
encouraged lots of consistent third-party software by having a huge amount of built-in
system software and heavily promoted design guidelines. But we kept the design
proprietary to ensure a virtual 100% market share of the GUI segment, which we had
for eight years. And even today, when people write that the Mac is easier to upgrade,
maintain, learn, and use than a clone, they hastily follow that with “but it’s only
available from Apple,” as if they do not see that the latter was the cause of the former.
Microsoft’s scheme, of course, was to come from nowhere in the applications
business to near-total domination in three years by shifting the system software from
DOS (which Lotus, Word Perfect, and Ashton-Tate dominated) to Windows (which
looked a lot like the platform that Microsoft applications dominated - namely, the
Macintosh).
These schemes were simple in concept, executed well, and made billions of
dollars for the companies involved. They didn’t have to be an ounce more intricate, or
have any sinister motives or actions. They weren’t even recognized as strategies at the
time, and were even called bad business decisions by some, but the results speak for
themselves.
And all the schemes cooked up by those companies afterwards have not come close
to matching the success of the originals. Maybe there’s a lesson there for companies
both big and small: keep your schemes simple, execute them doggedly, and don’t be any
more clever than you have to.