Now that Microsoft is IBM, it needs to avoid IBM’s big mistake

Last Updated on July 14, 2017 by Dave Farquhar

Whether Microsoft likes it or not, it’s turned into IBM. The biggest difference I see is that when Microsoft makes a mistake, it catches up with them much faster than the same mistake did to IBM.

But IBM’s biggest mistake was its adamant refusal to compete with itself. And that’s what Microsoft is going to have to avoid. Like Computerworld says, Apple says if you don’t compete with yourself, someone else will.

That mistake ultimately cost IBM the PC business. In 1985, Intel released its then-state-of-the-art 80386 CPU. IBM feared this 25 MHz powerhouse because it would keep pace with its lucrative minicomputers, but a PC based on a 386 was going to cost less than those minicomputers did. So IBM just ignored the chip, continuing to sell slightly underpowered 80286-based systems.

So what happened?

Compaq didn’t have a minicomputer business to worry about, so Compaq took a 286 clone and adapted it to use the new 386 chip. It didn’t even come close to using the full capabilities of the chip, but it was a lot faster than IBM’s 286 systems. It caught on, and it wasn’t long before the industry looked to Compaq for leadership rather than IBM.

And that minicomputer business? Well, people bought Compaq clones of the 386-based PC that IBM refused to make instead of IBM minicomputers. That experiment led to the Intel-based servers you still use every day today.

By the time IBM released its first 386s in late 1987, it might not have been too late. After all, Intel made a similar mistake when moving to 64-bit and AMD caught Intel off guard with its AMD64 architecture, but Intel swallowed its pride, cloned the AMD64 architecture, and recovered. But IBM’s mistake was to overcompensate by trying to make the machines proprietary and uncloneable. With that, IBM lost trust, and, well, I’ve told that story before. IBM’s decline was slow and painful, but ultimately, after trying everything, they exited the PC business. That idea was unthinkable in the 1980s.

How badly IBM is hurting today is a matter of debate, but there’s no question they’re not as big as they could have been.

A short time ago, Microsoft and Windows’ dominance seemed at least as unstoppable as IBM’s PC business ever was. But just like Big Blue became Big Black and Blue, Microsoft has to be feeling pretty black and blue itself right now.

Although their business models are different–Microsoft is primarily a software company, while IBM was primarily a hardware company at the time–they shared the same addiction to high prices and high profit margins. IBM was a vertically integrated behemoth, and when they changed product lines in 1987, they made 70% of what was under the hood, so the cost to manufacture it dropped. So what did IBM do? Raise prices, naturally.

Microsoft needs to seriously consider how many more copies of Windows it could sell if Windows cost half as much. Maybe half as much is too much. Because here’s the thing. You can build a surprisingly decent mini-PC now for well under $200. Motherboards based on the Intel Celeron 847 or AMD E-350 cost less than $80. Add 4 GB of RAM for around $40 and some kind of storage, throw it in a case that once held a discarded Pentium 4 or something, and you have a sub-$200 computer. Well, except a Windows license for it costs as much as the hardware.

What will happen is people will build the PC, then run Linux on it, or pirate a copy of Windows. But if a Windows license cost $40, people wouldn’t bother. It’s easier to just pony up 40 bucks than to try to defeat Windows activation.

These low-tier CPUs won’t set the world on fire, but they’ll run a web browser, e-mail client, word processor and spreadsheet just fine. They’re even better at video playback, making them ideal as home theater PCs. Although they run at low clock rates, their overall performance is comparable to the early Core 2 Duos, and they use a mere 18 watts of power. And you can stuff 16 GB of RAM in them if you want, which could be nice.

Microsoft should be encouraging PC makers to make $200 Windows Media Center PCs out of these boards. Why wouldn’t they sell? Here’s a capable home-theater machine that can run older Windows games too, and you can check your e-mail on it in a pinch.

Of course, some people would buy them and use them as regular PCs, and that might cut into PC sales. Others might use them as cheap game machines and that might cut into Xbox sales. That’s why it’s not happening. But now we’re back to competing with yourself. People aren’t buying PCs anyway right now, so why not try expanding the market into something new?

But since Microsoft doesn’t want to compete with itself, people just buy an Apple TV or Roku instead.

If you found this post informative or helpful, please share it!