In the late 80s, IBM attempted to reassert control over the PC hardware platform with the introduction of the PS/2 and its proprietary MicroChannel architecture. The cloners fought back, customers voted with their feet, the PS/2 initiative failed, and the era of open PC hardware continued and flourished. This was hugely beneficial for MSFT as a thousand PC OEMs bloomed, PC-based innovation surged and costs dropped, and MSFT software rode the wave of market expansion.
And it was great for end users. Not only because it drove system costs down, but it also created a rich market of add-on products — everyone could mix and match hardware to create their optimal system, whether they cared about cost or performance or maintainability or upgradability or whatever. Corporations could spec out and build standard low cost machines, enthusiasts could build super-tweaked machines, verticals could build out specialty machines, all on the same open hardware platform.
In the last 15 years, though, the market has shifted dramatically towards the laptop form factor. This shift has been a relative disaster for MSFT. The industry has moved away from an open hardware chassis with mix-and-match components, to closed tightly-engineered all-in-one machines. This shift has played to Apple’s strengths in design and integration and has negated many of the benefits of the PC ecosystem. The PC industry is still struggling to figure out how to regain design and profit momentum — Intel’s Ultrabook effort being the latest scheme. But the Ultrabook is just a direct response to the MacBook, it does nothing to recapture the open hardware experience of the 90s.
The open hardware community still exists in various forms, but is no longer focused on the PC platform and is not much of an asset for MSFT. Enthusiasts still build PCs, mostly for gaming — Maximum PC for instance has a good guide to components, Newegg is the place to buy. But this isn’t mainstream any more. The “maker” community is vibrant but is focused on other platforms largely — Arduino, the Kickstarter community, etc. The vibe and energy around open hardware is great, but it is no longer tied to the PC experience and is no longer an asset for MSFT.
MSFT has always been great at chasing taillights and is hard at work supporting the Ultrabook, competing with the Apple stores at retail, pushing Windows Phone, etc. But chasing Apple’s taillights results in products that are more and more like Apple’s — fully integrated hardware/software/services, a captive retail experience. MSFT has to do all this, the mainstream of the market is here, but there is nothing distinctive about the resultant products and experience. The Ultrabook/Windows/Microsoft Store products may equal the Apple experience, and may offer users a few more choices of hardware brands (does anyone care?), but the experience won’t stand out. Necessary work but not sufficient to recapture thought leadership in the market — at the end of the day, MSFT will be able to claim parity but no more than that.
If I was in a leadership role at MSFT, I’d invest in strategies to recreate the open hardware platform dynamic around the Windows platform. It is not obvious how to do so with the laptop and tablet as the mainstream platform, but I would spend $100s of millions trying. MSFT clearly has the cash to spend on new frontiers and new adventures, a couple hundred million on an effort to change the basis of competition in the PC market seems like a wise bet, even if it fails.
How about putting a “maker’s corner” in every retail store with modified cases and modified machines, maybe even workshops? Get the energy of the PC gaming community into the store, let people see this energy. How can the laptop design be modified to support add on hardware — super high speed optical expansion busses, wireless high speed expansion busses, novel expansion chassis ideas? Sifteo cubes are kind of cool, can this idea be used to provide hardware extensions to laptops? Are there other ways to “snap on” hardware to extend the laptop or tablet, using bluetooth or induction or other mechanisms? Can MSFT seed the maker community with funds or tools? Can MSFT embrace Arduino somehow, or Kickstarter? Could the PC be the hub for thousands of Arduino-based sensors and actuators and gadgets? These ideas are all admittedly poorly thought out, and I am not sure any one idea is right, or if any will work. Rich astutely observes that the days of an open hardware platform may be behind us permanently, just as the glory days of car tuning and hotrodding are behind us.
But I would spend a lot of money chasing after any idea that would move away from closed all-in-one hardware designs, and I would experiment with many ways to reinject open hardware dynamics back into the PC/tablet market. Ultrabook is not this — it is a fine and adequate taillight chaser, but it won’t shift competitive balance back in MSFT’s favor.
This is not the only reason for MSFT’s stagnation in the last decade, there are many other aspects to consider, but the dwindling of the open hardware ecosystem has been a loss of MSFT. For another take on Apple’s success against MSFT in the last decade, check out Rich’s analysis — the observations about vertical vs horizontal integration ring true.
(This post replicated in part from John’s personal blog)