One of the most interesting changes this week was Microsoft placing the head of their Surface hardware effort, Panos Panay, in charge of Windows. This change may seem trivial, but it effectively destroys the model that created Windows in the first place but should result in products that can better match the reliability and experience associated with Apple offerings. DOS and Windows broke the mold of how you developed and sold technology products separating key components that would have more traditionally come from the same vendor into separate elements and companies, allowing Microsoft to reach incredible sales volumes but at a clear cost to quality and security.
This one staffing change changes the Microsoft model into something closer to what Apple had when they licensed, which failed, and like what IBM had with OS/2, which also failed. But given I was engaged with both companies when that happened, I think I know how this could be done more successfully, and both Apple and IBM showcased that while their models failed, they did demonstrate higher relative quality and security than Windows did at the time.
That quality and security have become more critical, and I think I know now how to make this kind of change successfully now. I should add that Microsoft is also far better at partnering than either Apple or IBM were when their models failed, and this suggests, even without my suggested changes, the result should be a far less aggravating offering as a result.
Let’s talk about that this week.
Separated At Birth
When the IBM PC was first conceived, its conception was affected by two things. IBM’s consent decree, which forced them to separate the hardware and software efforts, and Apples growing success, which represented a potential threat to IBM’s continued dominance.
Instead of using the operating system the company had developed internally, they instead licensed DOS from Microsoft who brilliantly had acquired and then licensed it to IBM at cost. Microsoft was free to license it to others, creating firms like Compaq, Dell, Acer, and Asus and a market with an incredible ability to scale to eclipse Apple was born.
Having this separation between software and hardware was new; generally, products were and are build as a complete whole, and while Microsoft had done the productivity software for Apple, they did their OS. Distribution was also separate, and those focused PC firms arose with little cost to Microsoft, which gave them their massive competitive advantage of economies of scale.
However, the resulting product was significantly less reliable, less secure (even security software initially came from different companies), and the result was far lower customer satisfaction than Apple enjoyed offset by what was often a significantly lower price.
Microsoft has been undoing parts of their model over time, first with security, which significantly increased the security of the product, and they have worked at better coordinating with hardware developers, particularly concerning drivers and Windows 10 is night and day better than Windows 95 was as a result.
We’ve gone from hours of uptime to weeks, and even though we still get breaches and malware, the product is generally resisting even State level attacks now where before it had issues resisting kids’ efforts to create malware.
But it still couldn’t approach the perceived quality of an integrated product like the Mac even though the Mac couldn’t approach the same value for the dollar that Windows enjoyed.
Why Apple’s Effort and IBM’s Effort Failed
Both Apple and IBM failed with something like what Microsoft is attempting, but both companies were in very different places. Apple was a premium vendor, and what happened wasn’t that the licensees built bad products; it was that they built good far less expensive offerings that were more cost-competitive with their Windows peers. This move stripped massive amounts of revenue from Apple because their buyers weren’t premium buyers, they were just paying premium prices, but the added complexity did result in more breakage degrading the Apple premium brand. It was somewhat like what happened when Porsche brought out the VW powered 914 and 912. It cost them revenue and did brand damage so it couldn’t sustain. For IBM and OS/2, they didn’t have a critical mass of partners, the firms Microsoft had didn’t trust IBM, and the result was they had trouble even giving the product away (at one point it was put in cereal boxes as a free treat).
Microsoft has the critical mass that IBM lacked, and they aren’t predominantly a premium hardware vendor like Apple, so the causes for the failures in Apple and IBM don’t currently exist at Microsoft.
However, to make sure they don’t kill their PC OEMs, there should be some changes.
Benefits Of Change
By more tightly coupling hardware and software inside Microsoft, we should see more advances like Windows 10X, which will more aggressively take advantage of dual screen products like the Surface Duo, allowing the product to more aggressively innovating and driving excitement back into the platform. Besides, there should be even fewer opportunities for breakage, and the potential to provide an experience that exceeds Apple’s while holding the price/cost advantage they already enjoy should result. In short, we should see a more reliable, more attractive, and more innovative line of products emerge, but they’ll still need to protect the OEMs; otherwise, a competitor will emerge or advance (like Chromebooks) to fill the gap Microsoft inadvertently opens with this move.
Wrapping Up: 3 Changes To Assure Windows Future
To make this work, Panos Panay needs to be measured only on the success of the platform, Windows, in terms of sales volume and quality. On this last, I’d recommend NPS over any other quality metric. If he is measured on hardware sales cannibalizing the OEMs becomes attractive, and that will adversely impact Windows volume and success.
Second, the OEMs need to be brought in more aggressively and formally as peers to Surface, so they don’t feel like Surface has an advantage. Any unique advantage offered to Surface needs to be equally available to the OEMs, and decisions should be made on the advice of both internal and external groups with a focus on doing what is best rather than favoring either group over the other.
Finally, the focus of Surface on Apple and Google, rather than cannibalizing OEM revenues, needs to be reemphasized constantly. If the OEMs see Microsoft’s effort targeting them, they will likely shift away from the platform over time, favoring a vendor like Google, who appears more neutral by comparison.
With these three changes, I think Windows can be consistently be made stronger rather than trading off a market share for product quality and creating the opportunity for either a Google Chromebook surge or the emergence of another top tier competitor.
In the end, there is no doubt that this change will improve Windows quality; what is in question is whether it will have an adverse impact on Windows market share. That will depend on Microsoft’s ability to execute, and this decade, that ability is impressively strong.