|
by Commander Tansin A. Darcos 07/09/2013, 8:04am PDT |
|
 |
|
 |
|
When IBM upper Management saw how well other small companies were doing making microcomputers, they decided that IBM should do one that's competitive with others, but with the higher margins IBM can command because of their reputation for quality.
Only problem is, IBM is not a microcomputer company. They've only made one small computer, the 5150 that was actually an IBM 360 mainframe processor stuck in a small computer package. The built-in language on it was the most powerful computer programming language ever developed, APL. No wimpy Basic or even powerhouse C, no, they went with a nuclear-weapons grade language. If you're good at APL, you're about ten times as productive as anyone else using any other language ever developed. But this small computer was expensive and it was really just a mainframe with a master console, it wasn't really a PC the way we think of them.
IBM had quite a bit of success with its System/34, System/36 and System/38 minicomputers too. If your local supermarket has IBM cash registers, they were originally dumb terminals connected to a System/38 minicomputer. (Later after IBM developed PCs, they could now make their cash register systems into dumb terminal registers connected to a smart Personal Computer as the server machine.) IBM can do multi-terminal minicomputers very well, but it's not really a single-user machine company.
So, Don Estridge moved his team off-site, to a leased building in Boca Raton, FL. To be able to get something done in a reasonable time, he ordered off-the-shelf components from 3rd Party suppliers instead of using IBM's own manufacturing facilities. IBM makes lots of different processors, but Estridge's division went to Intel and got one of theirs, a hybrid 8/16 bit processor called the 8086.
To get the machine to be able to run, they needed an operating system. They went to Gary Killdall, who owned the company that makes CP/M; he was apparently too busy to see the representatives from IBM and so they went to another small company, named Microsoft. Microsoft didn't really have an operating system, so it bought one called QDOS (for "Quick and Dirty Operating system" and modified it to fit what IBM needed. They then developed a Basic interpreter that IBM could include in the ROM of the computer. IBM got the software needed to get the machine running as PC DOS; Microsoft got the right to merchandise it themselves as MS DOS. This was the rocket that propelled Microsoft to be the huge company it is today. Microsoft would later return the favor by screwing IBM over on OS/2, a regular practice they've continued ever since, anyone who partners with Microsoft is like a small animal bedding down with an elephant; eventually the elephant rolls over and crushes them.
To get actual applications developed for this machine, Estridge realized IBM could not do everything; he got IBM to encourage Third Party developers by publishing everything openly. And they developed an applications sponsorship program, in short, creating an Apps store back in 1980.
It was a big success because Don Estridge realized that the culture of IBM would not permit him to develop a microcomputer and he would have to build it outside the company (but within the umbrella of its resources and support). It seems that most successful companies have discovered this is how you remain successful, even companies like 3M, when someone wants to try to develop something, they get in-house support to work on it, but if it gets large enough, they move it to its own facilities, if necessary, it's own factory, and it, in essence, becomes its own fiefdom, they run things their way, and thus, they set things up to fit their environment.
The success of a number of companies who are world-class in their field came because large companies that are really successful realize that individuals can't really change bureaucracies, the company changes them, and they have found that to make things work, a large company really needs to be a lot of small companies under a management structure that can handle it.
One of the things reported in In Search of Excellence was that you can't run a factory or a department much bigger than about 500 people or so, and when you start to get much bigger than that, it's time to split the department into two or more separate ones. This seems to make a lot of sense, because the general manager of a department could have, say, ten managers reporting to him, and each manager having 50 people report to him, and thus you have only two levels of management and you can get to know, personally, 50 people. If you get a really good manager and a really small division, say 100 people, it's conceivable she or he could know all of them.
One of General Electric's engine manufacturing divisions which makes airplane jet engines has one manager, a woman, and 300 employees. And it's one of the best producers of this type of equipment in the industry. It's small, they don't try to do two or three things, and if they get an idea for something other than Jet Engines, they'll move the person who champions the concept and their project to another department or again, if it's big enough, to a separate facility all their own and it becomes independent of other departments and it thus can establish its own culture that works for them.
GE's previous CEO, Jack Welch, was well known as a manager who would 'scrounge for resources' in order to bootleg projects, and this is one thing company culture encourages. 3M does something similar. Did you know at one point 3M was making board games? Someone got an idea, they got funding, they went off and did it. I believe management realized that while it was successful, it doesn't really fit 3M's core competency, so the game departement was sold off to another company. But the point remains that 3M is one of those companies that realize if you're going to do something different, you can't make it work in an environment where the culture has crystallized already, it either has to fit in to a culture that will work with what its doing, or it needs the freedom to create its own culture, rituals, myths and standards.
That's why, for example, if you were going to run a farming company you have to know what is compatible. You can pretty much run a corn farm with the same rules as a wheat farm except for the differences. You would not dare do that if you were going to operate both sheep ranches and cattle ranches, you'd best run them as if they were separate companies, lest you end up with actual fighting. You could run a company where they had six cattle ranches or six sheep ranches run by integrated management. No sane person would seriously try to have the same person or the same standards used for managing three cattle ranches and three sheep ranches, the cultures do not like each other and will not integrate.
Valve hasn't learned this lesson, that if you are going to focus on something radically different, you cannot change the company culture. Instead, you move it outside the company culture and let it establish its own. Microsoft is really good on this - plus they never did wild-ass 'bet the company' schemes that killed so many of their competitors - and that's why, despite their screw ups - "Microsoft Bob" was their implementation of New Coke - they're still around, while the prairies of the Wild Wild West of Software Development are littered with the corpses of the failed companies that blew it. Or pulled stupid stunts equivalent to the Donner Party, and end up having to use the same solution to fix it: eat themselves.
|
|
 |
|
 |
|
|
|