Mitchel Lewis


Broken Windows

Just as our eyes are the window to our soul, Windows is the window to Microsoft’s soul and in spite of what their rising share price might indicate, Microsoft is destined to hit a plateau and stagnate if they maintain present course. Obviously, you would be a fool to deny that they are absolutely killing it with Azure and Office 365, but those products cannot stand on their own without Windows. Acting as both a lynchpin and as a funnel for the rest of Microsoft’s products, Windows creates a natural upgrade path for Office, Active Directory, Exchange, SQL, Sharepoint, Office 365, Azure, CRM, and others as businesses scale. Without Windows creating a natural upgrade path for these services though, they would be forced to run on merit alone which is something that they were never designed to do. Needless to say and just as it has always been, when Windows is in jeopardy Microsoft as a whole is also in jeopardy and things don’t seem to be going well for Windows these days.

I am not at all implying that Windows 10 isn’t doing a good job. Although Microsoft tries to hail it as the “last” version of Windows, it is a completely new operating system, especially when measured by how it generates revenue. Where Windows was previously built around a licensing model where OEMs, businesses, and retailers bought licenses in bulk to use and resell, Windows 10 was built around being free, at least for consumers, at the expense of telemetry, which cannot be disabled, and targeted ad revenue, similar to that of Google. Essentially serving two masters, it’s not really built with the sole focus of enabling business anymore, which seems to be why they are harping on their user counts these days much like any other search or social network built around ad revenue. As a wildly new approach to an operating system though and regardless of what anyone has to say about it, Windows 10 is a massive success in the free market and has NOTHING to be ashamed of as the #2 operating system in the world. Not only did it seem to soak up the lingering Windows XP, Vista, 7, 8, and 8.1 users like a sponge, but it did it while being free; at least for some time.


Sure, Windows 10 has a long way to go and still lives in the shadows of Windows 7, especially in the enterprise, but for a new OS that is free, it is a fair apology for wasting everyone’s time with Windows 8 and 8.1. However, it is by no means an upgrade in comparison to Windows 7 for commercial, industrial, business, and enterprise use which is why Windows 10 has failed to make a significant impact in this massive market sector, similar to 8 and 8.1. While they are notorious for hating change like everyone else, there is a logical reason to shy away from Windows 10, especially on an enterprise scale where individual virtues of software and hardware become much more prominent.

Although their ridiculous update frequency in Windows 10 has been well documented, few seem to take into account what such a frequency is truly indicative of, especially when looking at the types of updates that they are shipping. Microsoft isn’t exactly rushing out feature updates and Windows 10 had twice as many non-feature and security updates in 2016 as Windows 7 in 2015. This is not only symptomatic of lacking quality protocols that are unable to keep up with their sprints, essentially placing quantity above quality, but is also highly indicative of a drastic uptick in the defect density which also happens to be the logical consequence of such an approach. Simply put and to those unaware, an increased defect density tends to result in the uptick in non-feature updates that we are seeing in Windows 10. But this is not a guarantee in itself, correlation does not equal causation.

Source JNUC 2016:

While Microsoft’s QA practices or the defect density of their products are incredibly difficult to verify from the outside in, you can still measure how defective a solution is by it’s impact to business, especially in comparison to another or itself as time goes on, primarily by the impact to total cost of ownership (TCO). For example, IBM began rolling out Apple products to their users when they noticed that users on macOS reduced their TCO to 1/3 that of a Windows 7 users. To no surprise, when comparing the non-feature updates of Windows 7 and macOS of that same year, macOS was 1/4 that of Windows 7 in 2015. With this in mind though and when considering that Windows 10 had 270+ non-feature updates in 2016, more than twice as many non-feature updates as Windows 7 in 2015, then what can we expect from Windows 10 with regard to its impact on TCO? When combined with a significant rise in TCO is highly indicative of a steep rise in their rate of defect and defect density.

Hypothetically and when measured by updates alone, if macOS is generating 1/3 of the TCO of Windows 7 for IBM, a company that most would agree is incredibly efficient at managing PC’s by virtue of inventing them, then it should generate 1/6–1/8th of the TCO that Windows 10 would generate in the same environment. For those unaware, this means that a business could buy their employees redundant well-equipped MacBook Pros while still generating significantly less ownership costs than the same employees on a single Windows PC, even if the PCs were free. Although their initial costs are often compared, the costs businesses incur after purchase required to manage and support these products almost always dwarfs the initial cost.

IBM’s Mac Presentation @ JNUC 2016

When measured by TCO, what truly matters for businesses, Windows 7, a masterpiece in its own right and Microsoft’s best foot forward, gets flat out embarrassed by macOS as shown by IBM and Windows 10 is nowhere in sight; hence why business, industry, and enterprise are still holding out on Windows 7. And this isn’t just deferring to IBM because they’re a smart company, but because they literally wrote the book on the understanding that I’m applying. When viewed internally and when looking at complexity, we can see that software decays over time and becomes more costly as it becomes more complex, which is what measuring defect density allows us to correlate and this is exactly what is happening with Windows.


With this in mind and besides pure lunacy, what option besides Apple does business and enterprise have but to go with Apple when a fully loaded MacBook Pro costs significantly less than Windows PC after 3 years of use on average, even if the Windows PC were free? Where do they go in 2020 when Windows 7 runs out of official support and see an option of a drastically more efficient solution in macOS vs. a relative nightmare by persisting with Windows or migrating to Windows 10? Further and if businesses abandon Windows for macOS, then what would happen to Office 365, Azure, and other ancillaries without the natural path of adoption created by Windows? My guess is that they would be about as appealing as the notion of having the Edge browser on macOS. This is bad news for Windows and Microsoft.

Rather than addressing these problems and making Windows more competitive for businesses, specifically by code refactoring and making a focused effort on bugs, Microsoft seems to be more than willing to put lipstick on a pig and do some dirt in order to monetize it in other ways. For instance, by converting Windows into an ad platform that is retaining creepy amounts of telemetry data (which cannot presently be disabled), they bolstered Bing’s revenue significantly. Like any other ad platform though, they are desperately dependent on adoption and when Windows 10 didn’t get the reception that they had expected, Microsoft resorted to deploying their new OS via a recommended update that couldn’t be opted out of, stooping lower than the AskJeeves toolbar and resorting to malware tactics that have plagued their own OS for decades. They’re even goig so low as to instruct their own partners, business, and IT consultants, to “create stickiness” with their products as a means of entrenching them and increasing switching costs. None of which are the actions of a company whose products can run on merit, but that of a too big to be regulated monopoly scraping the barrel for relevance.

Regardless of their share price, these are all desperate acts if you’re a technology company such as Microsoft. Where they used to be able to charge hundreds of dollars for a license of Windows in 2009, they are having trouble giving it away for free in 2016. Where Windows used to empower people, it now preys on them like malware or Facebook; although not to the same degree or in a manner that is devoid of net benefit. Where they used to be THE platform, they now have to build their products and services to be agnostic of it today. Where Apple can sell a 3-year warranty and support contract directly to their customers for $200 which is less than a single Windows support case from Microsoft, Microsoft products create a multi-trillion dollar partner industry employing 17 million people worldwide. This is a significant amount of disparity and, again, bad news for Windows and Microsoft as whole.

Unfortunately and no differently than cars, acknowledging a defect problem is much easier than resolving one. When bugs make it into production, they can be 50–1000x more expensive to resolve in hindsight than if they were to be resolved prior to reaching production. Not only do you have to fix mistakes and redo a mountain of work, but you have to fix everything bound to and dependent on that mistake, which is a costly affair in an industry where your engineers drive Porsche’s. Needless to say, significant resources would need to be expelled on Windows 7 and/or an unprecedented amount of resources on Windows 10 in order to bring them on par with macOS and Microsoft would be MUCH better off breaking new ground because of this. Could this be why Windows 10 has been dubbed the last version of Windows?

Simply put and at scale in the enterprise, Microsoft hasn’t had a relevant OS since Windows 7, in almost a decade, and there is an obvious trend of businesses and enterprise such as IBM and GE aligning with the Apple ecosystem along with Google’s suite of alternatives and are finding a great deal of advantage in doing so. This is not by virtue of their products being trendy or because these companies are anti-Microsoft, but because macOS is an Abrams Tank in comparison to the musket that is Windows, at least from the perspective of an accountant or an engineer, and it’s only a matter of time before the rest of business and industry catches on to this fact. Microsoft probably realizes this though, hence why they’re calling it the last build of Windows and why they were at the aforemetioned JNUC conference trying to coerce Apple users onto Azure through JAMF. Or do you honestly think that Microsoft, the preeminent software engineering firm for decades, is ignorant to the ramifications of software complexity and defect density?


Because of this, Microsoft should have 2–3 years of leeway in order to do the impossible and bring Windows 10 up to snuff or build a new OS entirely, and/or bolster all of their other cloud services and revenue streams so that they can survive in a commercial realm, their chief source of revenue, which may not dominated by Windows. As stated before, their products are not designed to stand on their own, but to be ancillary to the ecosystem surrounding Windows. Microsoft is already addressing this by striving to be as accommodating and platform agnostic as ever with their cloud platforms which is wise a move regardless of their motivations, but will this be enough if Windows loses ground in 2020? Unless drastic changes are made to their operating system, it is my opinion that businesses and enterprises will have a fiduciary duty to follow in the footsteps of IBM and GE, take the path of least resistance in the direction of Apple and Google while moving away from the Microsoft ecosystem entirely as a consequence. All of which would be a crippling blow to Microsoft as Windows remains their #1 source of revenue despite it’s shortcomings.

More by Mitchel Lewis

Topics of interest

More Related Stories