Sign up to gain exclusive access to email subscriptions, event invitations, competitions, giveaways, and much more.
With big layoffs rumored to be close on the horizon, IBM is going to need to use another regeneration.
IBM CEO Ginni Rometty may be readying the ax
With massive layoffs rumored to be in the works at IBM, it looks like the storied company is about to go through a profound transformation. But Big Blue has seen this before, going through boom-bust cycles and different focuses. IBM has in fact tried on a lot of different identities over the past hundred-plus years. You could say it's been a lot of different companies over the years -- some of them overlapping, to be sure, but most of them quite distinct.
The tabulator company
Before there was IBM, there was Herman Hollerith, who invented a punch-card machine to help complete the 1890 U.S. Census. (The 1880 census had taken seven years to finish by hand.) In 1896, Hollerith founded the Tabulating Machine Company to market his creations, which quickly became a hit with government agencies and railroads. Hollerith's company merged with three others in 1911 to form the Computing Tabulating Recording Company. His tabulating machines, which revolutionized inventory control and accounting, are what's best remembered today, though industrial time clocks were also a great revenue source. In 1924, when longtime general manager Thomas Watson ascended to the chairman position, he changed the company's name to International Business Machines. The last IBM tabulator, the IBM 407 Accounting Machine, was sold until 1976.
The mechanical office machine company
What is a "business machine," anyway? Could be just about anything IBM ever made, of course, but few things seem to epitomize the phrase more than devices like the IBM 801 Bank Proof Machine, a furniture-sized piece of equipment that made it so much easier for banks to process personal checks that the company sold hundreds of them right into the teeth of the Depression. Though in retrospect these machines are the obvious predecessors of computers, their specialized functions and mechanical nature really put them into their own now-outdated class.
The defense contractor
When World War II broke out, IBM transformed its factories in the United States so that they would produce war material, including bombsights, rifles, and engine parts. During the war itself, these were sold at a nominal 1% profit, but the company would go on to have a solid business relationship with the defense establishment as the Cold War developed. For instance, the company built the SAGE system, which was intended to detect any intrusion into U.S. airspace.
The mainframe king
A thousand years from now, this is what IBM will be remembered for: the huge, powerful computers that started in the 1950s with the IBM 701 (originally known, as the previous slide should've helped you guess, as the Defense Calculator), hit their stride in the '60s with the multipurpose System 360, and continue on today with the System z.
But the mainframes attracted a 1969 antitrust lawsuit that changed the company -- and the industry -- forever. Though the suit was eventually dropped in 1982, the company had already reacted by unbundling the hardware, OSes, software, and services that had previously been packaged together in mainframe sales, thus creating several new markets at once. Still, Michael Frishberg, who worked at IBM in the '80s, told me that even when IBM PCs were selling like hotcakes, the mainframe division hadn't lost its internal prestige, and had its own special locked floor in the building where he worked.
The typewriter company
In 1933, IBM bought Electromatic Typewriters Inc., mainly for access to the company's patents and technical expertise; nevertheless, the first electric typerwriter released under the IBM brand became a hit, and the product line became a solid moneymaker for the company. In 1961, IBM introduced its Selectric line, with its distinctive golf-ball-shaped head, which became the de facto standard typewriter and an early input terminal for IBM's computers. IBM computer scientist Bob Bemer, developer of ASCII, tried to convince the Selectric group to implement the new standard in their typewriter, but the idea was brushed aside -- an episode emblematic of the turf wars that marked IBM's history.
The PC maker
IBM mostly missed the minicomputer boom of the '70s, but shocked the world in 1981 when it unveiled its own microcomputer, which it dubbed the PC. Developed in secret over less than a year, and built almost entirely out of commodity parts so as to be affordable, the product line in only a few years blew past Apple, Atari, and the other companies that had dominated the nascent market. (Still, Diane Piper, who worked there at the time, says, "As popular as they were, the managers 'in house' wanted nothing to do with them. It was a new animal, and you know what change is like in any environment.")
The PC's commodity components made it easy to imitate, and by 1988 clones were outselling IBM products by 50%; a falling out with Microsoft over OS/2 meant that IBM plans to dominate the software side also faded. The ThinkPad laptop line maintained its cachet well into the '00s, but even that was eventually sold off to Lenovo.
The semiconductor manufacturer
One of the most fateful decisions made by the PC developers was to use Intel's chips instead of IBM's own in-house semiconductor knowhow. Former IBMer Frishberg told me he still thinks IBM's chips could've been more widespread if the company had committed sooner. As it was, IBM silicon didn't find its way into mass-market PCs until the '90s -- and when it did, it was in the form of PowerPC chips designed in an alliance with Motorola and destined for computers made by Apple, IBM's erstwhile arch-rival. IBM never caught up with Intel, and even Apple jettisoned PowerPC in the mid '00s. IBM still uses its Power chips in specialized systems, but no longer builds them -- it sold off its chip manufacturing facilities in 2014.
The enterprise software company
In the mid '90s and early '00s, IBM had something new to sell to its business customers: not just machines, but the complicated stuff that ran on it. Microsoft could have the easy stuff like office suites: IBM would sell higher-end (and, hopefully, higher-profit) software -- some developed in-house, like its WebSphere middleware, others acquired, like Lotus Software (hands up if you've ever torn your hair out using Notes) or Rational's development tools. All this stuff was meant to be bundled up into "solutions" that could be sold to customers as part of larger contracts that would include lucrative ongoing maintenance.
By the 1990s the company had pushed its energy full-on into "services" mode, essentially becoming in many ways a team of consultants (which was only cemented when IBM bought PriceWaterhouseCooper in 2002). Jon Russell, who worked on IBM's x86 servers in the '00s, explained how this could get confusing: "You want some dedicated hardware? How about IBM BladeCenter? How about buying IBM software to manage your BladeCenter? The folks setting up solutions for the customer seemed to have dual roles: (a) find the best solution possible for the customers to have their needs met, and (b) have that solution be an IBM product. So my role in this is, if the customer goes with IBM, and if IBM finds they need hardware, and if they sell IBM hardware to the customer, they might also sell them the product I worked on. But maybe not."
The bean counters
The company has always been frugal -- Diane Piper told me about her time in '80s in the Suggestions Department, which fielded internal ideas for greater efficiency and rewarded the ones that worked. But services-era IBM became more and more focused on cutting costs. One former delivery project executive described to me his travails trying to provide good service to clients: coming in under budget in one area of a client contract and having to record that as profit rather than shifting it to another area the client needed, for instance, or having to wring profits out of contracts that had originally been lowballed and meant to serve as loss leaders. (Naturally, delivery project executives' pay was tied in part to customer satisfaction.)
The (sometimes misunderstood) prophet
Still, for all the complaints I heard from former IBMers, they also spoke warmly of their time there, feeling like they really were on the cutting edge of what was technologically possible. Eileen Mullen and Michele Tepper both worked on the IBM.com site in the heady early days of the Web in the late '90s. Tepper says IBM "was early in recognizing that the Internet would become a utility service, but a little slower, as large companies often are, to be able to act on it." Mullen's group got to help power websites for other companies making their first Web forays. And Michael Frishberg said he did work in the early '80s that could've led to modern cloud computing, which after all isn't that different from the mainframe time-sharing model. (Utility computing, an early '00s IBM buzzword, is a more obvious predecessor.)
The singularity herald
IBM is now trying to find a way into the future. For instance, the company has had an interest in artificial intelligence since Deep Blue took on chess champ Gary Kasparov in the late '90s; the company built Watson, a "cognitive computer" that successfully took on Jeopardy! champs, in 2011, and now its smarts are being put to practical use. So maybe IBM won't be remembered for its mainframes a thousand years from now. Maybe it'll be remembered as the company that made humanity obsolete. Stay tuned!