When You Use Windows 11, Are You the Product? (Premium)

Former Microsoft engineer Dave Plummer posted an interesting video this past week about Microsoft’s shift to free Windows upgrades. He explains why he thinks this happened–declining PC sales, the Windows 8 backlash, Linux, and Apple’s push to make macOS upgrades free are key among them–and what he sees as the “hidden costs” of this change: Forced telemetry, advertising, and a shift to subscriptions.

Sound familiar?

It should. We’ve been discussing these topics here since, well, the beginning. And while Plummer doesn’t provide any new information here–not for the first time, unfortunately–his videos are so widely viewed that the issues we’ve all been debating for years likely feel new for much of his audience. And to be fair, there is a discussion to be had on this topic. Windows is always top of mind (for me), and in this era of enshittification, I feel the quality problems in Windows 11 in an almost visceral way. These problems so serious that I’ve made major shifts to non-Microsoft solutions for cloud storage and writing that would have been inconceivable just a few years ago.

But before diving into this further, we need to discuss Dave Plummer.

?‍♂️ The problem with Plummer

As I wrote recently in Ask Paul, I subscribe to Dave’s channel and watch every video he produces. He’s interesting, well-informed, and technical, and thanks to his history at Microsoft, he is by definition someone I need to pay attention to. As the author of a book about the history of Windows and someone who has been covering this operating system professionally for over 30 years, I was curious–OK, I was excited–when he posted a video about Longhorn recently, assuming he’d have some nuggets of new information that even I hadn’t heard before. His epic interview with NT architect David Cutler is essentially viewing/listening, after all. But there was no new info.

This was surprising and disappointing. I realized while watching it–and then rewatching it–that a key reason was that Plummer wasn’t actually at Microsoft for most of that saga. He went on sabbatical before Microsoft revealed Longhorn to the world at PDC 2003, and then he left the company for good instead of returning. He wasn’t there for Longhorn, and he didn’t have any internal information that might have put that video over the top.

Well, Plummer wasn’t at Microsoft for Windows 10 or 11 either. And while he must still know a handful of people still working at the software giant, and possibly even some involved with Windows, I had better and more contacts within Windows and Microsoft during this time than he did. And I know a lot about the internal machinations that led to key strategy decisions with Windows 10, in particular. I was very close with Terry Myerson, for example, and most obviously.

This isn’t a contest. As noted, I’m a fan of Plummer and his YouTube channel, I respect him on a technical level, and I look forward to every video. But as he gets into topics that I’m intimately familiar with, it’s clear when he’s not on the same footing. And with the Longhorn video, and now this latest video, there’s just nothing new there. So that’s a bit disappointing.

? The problem with Windows 10

I covered the creation of Windows 10 in real time, mostly here on Thurrott.com: One of my last work trips during my many years at the organization that became Windows IT Pro–where I was most closely associated with the SuperSite for Windows, which I created back in 1998–was to San Francisco in late September 2014 for the Windows 10 unveiling. This was in the wake of the Windows 8, Windows RT, and Surface debacles, Microsoft’s subsequent apology tour with Windows 8.1+, and a major internal reorganization in which the team responsible for Windows Phone took over “Big” Windows.

The stakes were high. As Plummer notes, Satya Nadella had become Microsoft’s CEO that past February. As Plummer doesn’t note–it’s bizarre that he never mentions this topic at all–we were barely two years into Windows Phone, a platform Microsoft was about to change fundamentally for the third time so that it would finally align with “Big” Windows. And as Plummer notes, but gets wrong, “Microsoft needed to adapt from a Windows-first mentality to a world where cloud services and ongoing customer engagement mattered more.” That’s not what was happening in 2014. The “Windows-first” strategy had died decades earlier, replaced by a “Windows best” strategy in the early 2000s as Office and then Server had become more important than Windows and, with Office, more lucrative too.

What he’s alluding to there is now well understood: Nadella took the cloud-first strategy that had started under Steve Ballmer and formalized it into a dogma that all product groups within Microsoft had to follow. Literally. As with many new CEOs, he met with the leaders of each product group early on and made two demands. They had to justify their products as standalone businesses that were or could be profitable on their own. And these products had to make sense in the cloud-first Microsoft.

For some products groups, this was easily: Office had transitioned to Office 365 and then Microsoft 365 as a poster child of sorts for what the new Microsoft could be. Phil Spencer convinced Nadella that Xbox could make sense as a cloud-first business, which is what led to Game Pass, Cloud Gaming, and the Activision Blizzard acquisition. And Windows … well. That didn’t go so well.

Windows is by definition a client OS that sits between apps and services and the underlying hardware. There are two sides to this product, business and consumer, and the business side of the group was pushed into Microsoft 365 because those customers were already paying for the product–really, for support–via what we now call a subscription model, and bundling was a proven revenue earner, with upsell opportunities. Consumers were always the problem. They never paid for upgrades at retail to begin with–Windows upgrades were still correctly seen as unreliable and technically challenging–and they were never going to “subscribe” to Windows.

There were other challenges, of course. Windows 8 had landed with a thud, thanks to the customer base rejecting its touch-first interfaces and PC makers rejecting tablet and other touch-first form factors. So one of the goals for its successor was to completely roll back that mistake and return to a desktop-centric interface that customers could embrace. Another was the alignment with Windows Phone, which would become Windows Mobile again, on multiple levels: Architecturally, with the mobile apps platform, and from a look and feel perspective.

But Windows also had to make sense within the new cloud-first Microsoft, somehow. And here, the new Windows team took on what seemed like an impossible challenge. It would update this legacy codebase, this archeological dig of different technologies, many dating back decades, as if it were a cloud service. Myerson and his group called this initiative Windows as a Service (WaaS). We didn’t learn about that–or that Windows 10 would be a free upgrade across Windows 7/8.x PCs and very recent Windows Phone 8-based phones–until January 2015, at Microsoft’s second major public event for Windows 10.

The timing of this was interesting for me, as I had left Windows IT Pro/Penton at the end of 2014 and created Thurrott.com with George Coll and BWW Media. And so this was my first work trip for Thurrott.com.

Part of the shift to a cloud-first mentality–which, we should remember, was briefly “mobile first, cloud first”–required Microsoft to meet their customers where they were. This is why Microsoft Office finally came to the iPhone and then iPad and Android after Nadella took over. It’s why Xbox has expanded its view of the platform well beyond consoles. And it’s why Windows–and at the time, Windows Phone–were less important to the company. Windows was one place to run Office. It might even be the best place. But it wasn’t going to be the most common place. Mobile–in the form of iPhone and Android–had taken over personal computing. PCs were still important, and they still offered unique advantages over other form factors. But Microsoft’s customers, like the rest of the world, were increasingly consuming personal computing services on phones. The world had changed.

So we might summarize the transition from Windows 8.x to Windows 10 as a course correction. Windows would never again be given so much attention from senior leadership, nor would it factor in so highly in the company’s overall strategy. But Microsoft couldn’t afford to lose that audience–across businesses and consumers–and so the high-level goal was simply to stop the bleeding. Give those customers with PCs a reason to stick with PCs.

Looking back at the Windows team of this era, I still love their almost innocent, wide-eyed approach to the product, and that they were on a mission to set things right. But I also see the mistakes, how their naivety and inexperience positioned them for some rocky times. The shared app platform they created was mobile not desktop, and it would run on Windows PCs and tablets, but also phones, HoloLens, Surface Hub, Xbox, and other devices, and most of those would either fail immediately or never amount to much. And Windows as a Service? Amazingly, they eventually got that right, but there was a glaring, horrible mistake at the center of that strategy, and Microsoft wouldn’t fix it for years.

As publicly stated, Microsoft wanted to move as many customers as possible to one version of Windows, the latest version of Windows, called Windows 10. This makes sense on so many levels, but the biggest advantage there, something Microsoft saw Apple and, to a lesser degree, Google enjoy with their hugely successful mobile platforms, was that it was easier to support. If there were always two or three supported versions of Window out in the world, Microsoft would need to create different security fixes for each every time there was a vulnerability. Modernizing the user base would benefit everyone.

Unfortunately, that’s not what WaaS accomplished. Instead, Microsoft set out to release a major new version of Windows 10 every six months. Each Windows 10 version was accompanied by platform advances that were unique to those (and newer) versions, including the app platform. And so it ironically created a world in which there were more supported versions of Windows out in the world, not less. Plummer never mentions this problem in his video, it’s unclear if he’s even aware of it, and perhaps this is a separate topic. But it’s related, and important: WaaS was designed to help Microsoft keep Windows more secure, but it initially made that more difficult.

There are so many stories about this time, some apocryphal. For example, many seem to “know” that Terry Myerson “fired” the team responsible for testing and supporting Windows internally, and the commonly held rationale for that decision was that Microsoft now had the Windows Insider Program–revealed at the same September 2014 event at which it debuted Windows 10–and could thus use customers as guinea pigs. That’s overly simplistic and/or untrue: The Insider Program was an overt refutation of the Steven Sinofsky regime and its secretive, feedback-averse culture. Myerson wanted its biggest fans to be more engaged and feel like they were contributing to the creation and improvement of the product they cared about so much.

It took a few years–OK, it took several years–but Microsoft eventually course-corrected on WaaS. It calmed down on major new updates every six months, shifting to a slightly more reasonable major/minor cadence. As important, it stopped tying platform advances, including those related to the app platform–which by this point didn’t have Windows Phone/Mobile as a target anymore and saw basically zero engagement for non-PC usage–to individual Windows 10 versions. Going forward, new Windows features would work with all supported Windows 10 versions. (This is when the modern Windows platform shifted to the Windows App SDK and WinUI 3.)

Problem solved, right?

? The problem with Windows 11

Terry Myerson didn’t survive long enough to correct his mistakes and Joe Belfiore, his trusted lieutenant, came and went and then just went over a period of years. Windows was leader-less for a few years, with no representation in Microsoft’s senior leadership team, a first. And then Nadella inexplicably elevated a shyster into a new role leading both Windows and Surface.

The result was Windows 11.

Windows 11 was to Windows 10 what Windows Mobile 6.5 was to Windows Mobile 6: A surface-level spit shine with no foundational improvements and many user experience inconsistencies. But it was also worse, as its simpler new high-level UIs also came with functional regressions that triggered visceral, negative reactions from the very experts who by then were championing Windows 10. It was pretty but shallow, the dumbest part of the aborted Windows 10X effort that sought to bring reliability to the platform via a container-based architecture that would seal off unsafe Win32 desktop apps from the rest of the system. Windows 10X was apparently a “hard computer science problem,” as we say, and so Microsoft abandoned that effort.

Windows as a Service became “continuous innovation,” a seemingly vapid and innocuous term that many took to mean nothing at all. But we now see this as the threat–and the lie–that it was. Under the continuous innovation model, Microsoft shifted from two version releases each year to just one, pleasing just about everybody. But the asterisks to the conversation, the fly in this soup, if you will, is that the Windows team reserved the right to update Windows at any time. Instead of waiting for the big, annual release, it could simply add new features whenever it wanted.

And that’s exactly what it did. Worse, it escalated this behavior to the point we’re at now. Instead of including Copilot and about 1,000 other new features in Windows 11 version 23H2 in late 2023, it shipped all of those things a month early, via what seemed like yet another innocuous monthly cumulative update for 22H2. And it did that to force its corporate customers–who would have skipped 23H2 en masse because of all the changes–to adopt Copilot and the other changes.

Windows 11 was already enshittified, and it was always enshittified. But this is the moment when Microsoft took that to, ahem, 11. This is when chaos took over and when common sense went on its own sabbatical, never to return. This is when Microsoft changed Copilot repeatedly over several months, among other things, moving its Taskbar icon from place to place like a pinball. This is when Microsoft corporate aims were laid bare as being far more important to it than whatever Windows 11 users wanted.

Yes, this all started in Windows 10. The telemetry, which was the implicit agreement users made–most without understanding what it all meant–to keep Windows 10 safe, reliable, and up-to-date. You can have Windows for free, but you need to send Microsoft usage information so it can help ferret out problems that are perhaps isolated to specific configurations. But Windows 10 also escalated the inclusion of ads in Windows; this started in Windows 8, in apps, but Windows 10 brought ads to the OS. And Windows 10 is when Microsoft started mimicking PC makers by bundling crapware in the operating system, a repudiation of the quality previous generations of Microsoft leaders tried to bring to the platform via the Signature program and then Surface.

Windows 11 escalated all that behavior. And then escalated it again. And when Nadella required the Windows team to accept further cuts in both funding and manpower, issues that Myerson nad faced as well, its leader quit. And went to Amazon, notoriously the cheapest Big Tech company, and a place where funding and manpower are always problematic. What he left behind, again, was chaos.

?Look in the mirror, do you see the product?

Because of my focus on and affinity for Windows, I lived through these changes as they happened. I revisited this history many times. I wrote the article series that became Windows Everywhere over many years. I have read, recommended, and now re-read Steven Sinofsky’s Hardcore Software more times than I can count, and I am right now re-reading a series of Microsoft history books–Hard Drive, Gates, Over Drive, and several others–as I do from time to time. This is my life. I don’t get everything right, but experience changes you. And there’s always new information.

I did get a few things right.

When Microsoft put ads in apps in Windows 8, I called it a slippery slope, noting that this “cheapened” Windows and, more important, predicting–correctly–that this would only lead to more ads. And when Terry Myerson predicted that there would be over one billion Window 10 devices within a few years, I explained why that was impossible and then outed a Microsoft apologist for being the Mouth of Sauron in delivering the news of that failure in the lamest way imaginable. (Hitting that figure took twice as long as Myerson predicted.) And I called bullshit on the faux outrage over the non-existent problems with Recall, later pointing out that Microsoft made no substantive technical changes to the feature while engaging in PR theater and making very important non-technical changes. That Recall is a non-event in real-world use is almost beside the point. But maybe that is the point.

OK, even a broken clock is right sometimes, and when it comes to Windows and enshittification, I am absolutely beating the same tired drum again and again. But when I look at Windows 11 today, I see both sides of this product. I see the modern evolution of the platform and the security advances we’re now getting through Windows Hello Enhanced Sign-In Security and the other technologies that are now part of the so-called Windows Resiliency Initiative. And I also see the crap for what it is. The chaos of continuous innovation, the forced telemetry, the preinstalled crapware, the ads, and the harassment, not just to use Microsoft’s products and services, but also to use them in the configurations that benefit Microsoft the most while invading your privacy the most.

That brings us back to Dave Plummer and his recent video. And you thought I wasn’t going to bring this one home.

“Upgrading from Windows 7 to Windows 8 had cost $100 to $200 before,” he says in the video, though I will remind everyone that Windows 8 Pro was a $40 upgrade for the first several months and that the unsuitable touch-first user interface was a bigger problem than the cost. (And let’s not forget the very real issues I noted above about these upgrades being so difficult that most consumers never bothered regardless.) Windows 10 being a free upgrade was “a radical departure from the past,” Plummer adds, though I feel like we were always heading in this direction.

Windows 10 was a solid upgrade. It brought back the desktop focus, pleasing most customers. It corrected most of the mistakes from the Windows 8 era, though pushing forward with its less capable mobile apps platform was less defensible. Plummer does mention how the WaaS strategy “aligned with Nadella’s broader push towards cloud services and subscriptions,” which is correct. And that Windows 10 had become a “gateway to Microsoft’s ecosystem rather than a one-off sale.” which is a great description.

But are we the product?

The phrase, “when you don’t pay for the product, you are the product,” is so commonly said and believed that many have perhaps stopped even thinking about it. The most obvious proof point for this line of thinking is a product like Google Maps. We all seem to understand on some level that data mining is inherent to this incredibly useful product, and many, when confronted by the privacy implications, just accept that as inevitable if not a fair “payment” for the functionality. We get it, Google is tracking us. It’s selling that data to advertisers. Many of which we can see, clearly called out, right there in Google Maps. But we accept it, either willingly or begrudgingly.

These things exist. Indeed, they’re all too common. But I can come up with all kinds of personal technology products and services that I don’t pay for and don’t feel that I am somehow “the product” because the services are somehow stealing my personal information for some ill-gotten gain. Notion is a great example, and so is Substack, which my wife and I use for Eternal Spring. I don’t pay for either service and I am absolutely not the product. In Notion’s case, I have no idea why they don’t just charge me, as I’d be happy to pay. And with Substack, the product is Substack: The big complaint there is that customers are promoting Substack as much as they’re promoting themselves. Yeah, I guess so. But we both think it’s an incredible service.

But what about Windows 11? I noted the two sides of this product above. In what ways are Windows users … the product? How is Microsoft taking advantage of us in Windows 11?

Plummer notes that Windows is a where Microsoft can upsell customers to “things like OneDrive cloud storage, Skype calling plans, Xbox services, and Office 365 subscriptions.” Leaving aside the fact that Microsoft is literally retiring Skype in less than two weeks in another hint at how out of touch Plummer is these days with the company, fair enough. Microsoft does upsell customers on OneDrive, Xbox, and Microsoft 365 in Windows 11. Is this underhanded? No. Is it somewhat understandable, a fair exchange for this free OS we all use? That’s a conversation or debate, and I could see one landing on either side. I have argued for a paid version of the product that would remove these annoyances.

“The free consumer upgrade to Windows 10 was not a direct subscription,” Plummer says, “but it acclimated customers to the idea that Windows would continuously update … It was a clear signal that Microsoft envisioned the Windows of the future not as the $100 box but as part of a service-based model going forward.”

He overstates the use of telemetry in Windows 10–and I’m not sure what he means by “Windows 10, especially right at launch, collected an unprecedented amount of usage data by default.” This suggests that Microsoft somehow responded to negative feedback by collecting less data at some later point. I coined the phrase “privacy theater” at the time to explain that Microsoft made no substantive changes to Windows 10 data collection, opting to simply explain what it was doing and put up pointless new UIs in Windows 10. (Recall what it, ahem, did with Recall.) I always try to fall back on common sense, and I’ve often argued that we have enough real outrage to deal with, so there’s no reason to make things up.

But he calls Windows 10 (and thus 11) a “feedback machine feeding Microsoft constant insights about you and your computer use.” That’s alarmist and inaccurate unless you think the company’s privacy policy is an outright lie and that the data we send by default isn’t anonymized and used expressly to improve the product for everyone. If Microsoft isn’t lying, this is a fair tradeoff, a reasonable ask. But I do agree that anyone concerned with privacy should be able to opt out of telemetry. And Microsoft does not allow that.

Plummer does get something very right when it comes to telemetry, however, and this speaks to my comments above about those apocryphal stories so many just believe.

“By making the upgrade free, Microsoft ensured a critical mass of users would generate those [feedback] insights, allowing the company to refine Windows faster and more efficiently than in the old days of lengthy beta tests. This was part of a larger industry trend analogous to how Google refines Android all the time or even how Tesla updates their car software based on telemetry”

This is undoubtably true, and it touches on the real reasons behind the telemetry. Underlying a lot of what Microsoft has done with Windows since Windows 8 is this notion of making the platform more modern by copying what Apple and Google do with their mobile platforms. There are some advantages to that model.

But I can’t get over this commingling of what some believe with what is objectively the truth.

Plummer says that “Microsoft insisted that the data was anonymized and used just to improve the user experience” but that “the sheer scale of telemetry in a free OS made a lot of people uneasy.” OK. Well, that’s why Microsoft “insisted” that the privacy policy it created, which constitutes a legal agreement, is all the explanation it needs. We live in an age of misinformation, and countering a corporate policy with people feeling uneasy isn’t just unfair, it’s illogical.

“If you’re not paying for the product, you probably are the product,” he says. “And that classic adage was invoked by skeptics who suspected that Microsoft’s real strategy was to monetize Windows 10 users through their data and attention. Indeed, Windows 10 introduced a new wave of built-in advertising and upsell mechanisms within the OS itself. Users began noticing things like suggested apps and the Start menu and Taskbar pop-ups nudging them to try Microsoft Edge or OneDrive. And even full screen prompts to subscribe to Office 365. Microsoft wasn’t charging for Windows 10 up front, but the OS contained numerous hooks to drive users forward and into Microsoft’s paid services or their partner apps. In Windows 10 and, later, Windows 11, the experience out of the box includes these promotional messages.”

Yes. Yes, it does. We’ve been discussing all this for years. But are we the product? Or is this all a fair trade-off? Isn’t this what Apple and Google do, if less overtly? Does the existence of many workarounds and utilities aimed at lessening these annoyances in Windows factor into this discussion at all? (Plummer does mention that.)

Yes and no. Plummer brings up a good point about the app store in Windows, noting that it’s a “new revenue stream” for Microsoft because there are paid offering in there. And yes, it sure is. But this is overt and appropriate. You don’t have to use the store, and you don’t have to buy apps or games or whatever other content is in there. That seems like a normal relationship point between a company and its customers. It’s fine, and I like that he discusses it.

Giving away Windows did make sense given the climate of the day, just as it makes sense now. We accept these updates as easily as we ignored paying for or upgrading in the past. They’re just a part of the experience, a once a month (hopefully) occurrence. People use PCs less now than before, and they upgrade PCs even less frequently. Most of Microsoft’s customers–and by revenue this is even more important–are businesses, not consumers. As Richard likes to tell me when things go wrong, “This is how you know you’re not the priority.” That’s a reality that I–and, I suspect, many other enthusiasts–have been wrestling with for years.

OK, fine. We’re not the priority. But are we the product?

? Yea or nay

Plummer argues yes. He is among those who don’t want telemetry or advertising in Windows (or, more specifically, the “Pro versions”). Like me, he would gladly pay to be rid of this crap. He even goes so far as to say that Microsoft’s policies with Windows–its enshittification–have resulted in “a segment of users openly contemplated switching to Linux,” though there is basically no data to suggest that’s ever happened.

“The more people using your platform the more ways you have to earn money indirectly from that platform,” he says. “The economic landscape around Windows 10’s launch made a one-time OS sale much less important than the lifetime value of a Windows customer.” In short, Microsoft removed cost as a reason for a customer to leave Windows for macOS, Linux, ChromeOS, or some mobile platform.

In the end, this discussion is more nuanced than I’d like. So I have to again return to my original point about seeing both sides of Windows 11, which means that I’m clear-eyed about its pros and cons. In my case, the cons do not yet outweigh the pros, and I still very much prefer Windows. That said, I have been slowly shifting away from Microsoft offerings, not out of some plan or strategy, but because they no longer meet my needs or when they offend my sensibilities. This just feels like common sense to me. But you may land in a different place.

I respect that, just as I respect Plummer’s stance on this issue. But it’s also true that if the Mac, or Linux, or whatever other platform somehow improved to the point where using Windows was untenable, I’d make that switch. This is something I keep coming back to: I constantly try these alternatives, and I always come back to Windows. It’s important to point out that this is as much about the failings of those platforms–for me–as it is about the advantages of Windows. The reason I keep trying–whether it’s the iPad, Linux, the Mac, or whatever else–is that I’m a seeker, and I always want to work as efficiently as possible. For now, at least, those other platforms fall short, at least for my needs.

If I ever do leave Windows, or use Windows less often, or whatever, I hope it’s because an alternative improves and not because I’ve become the product. The slippery slope thing applies here, of course: It’s only going to keep getting worse. That makes me sad, but it means I also want to be ready should that day ever come.

Meanwhile, Windows is as Windows does. In some ways, it’s wonderful. In others, not so much. It’s not all I think about. But I do think about it a lot.