Windows 11 was originally released in October 2021, and a second new release (22H2) came out in December 2022. Overall, how is it doing? From all signals, it appears to be doing fairly poorly. One example, captured from Statcounter:
This measures usage based on web browser data — maybe not the most reliable way to measure usage, but it’s better than nothing (and Microsoft isn’t going to release any data itself). So what does this data show? Windows 11 — a free upgrade for anyone on Windows 10 — is slowly gaining ground, with Windows 10 numbers moving down by about as much as Windows 11 is going up (as you would expect). But it took Windows 11 nearly nine months to pass Windows 7, an OS that isn’t even supported any more (for most people at least). And that’s with new PCs being shipped with Windows 11 preinstalled. (At least I haven’t seen PCs being sold with Windows 10 preinstalled any more, probably because Microsoft won’t let the OEMs do that.)
So what went wrong? That part is a little harder to answer. Let’s look at a few possibilities.
Changed hardware requirements?
Windows 11 made some significant changes to the hardware requirements needed to run the OS. The most significant of those is the requirement for an 8th Generation CPU, meaning it supports machines released in late 2017 and later. For consumers, that’s probably a bigger deal since they keep PCs longer. For enterprises, the lifecycle for PCs is often 4-5 years, so that’s not really that bad. (And if enterprises were purchasing 7th generation or earlier previous-generation PCs past the end of 2018, they’re doing their employees a disservice.)
So what if organizations find that they have, say, 25% of their devices that still have older-generation PCs? Well, most don’t want to support two OS versions for more time than they have to, so they’ll probably stick with Windows 10 until that 25% are replaced with newer devices. So that would hold back the percentages.
What about the TPM requirement? Well, the OEM requirements for Windows 10 started requiring TPMs even earlier, so there’s no reason devices that meet the CPU requirement wouldn’t also support the TPM requirement. The biggest problem here is usually enabling those TPMs, as they may be turned off in the firmware and not accessible to Windows. But they are there.
How about the lack of an 32-bit (x86) Windows 11 release? This change was past due — everyone had already moved on to 64-bit machines. (That’s a fine way to make a change: wait until no one cares any more. But that certainly drags things out longer than necessary.)
Windows as a service changes?
Everyone wanted less frequent Windows releases, and with the announcement of Windows 11, Microsoft finally listened, with Windows 11 releases planned for once per year. Perfect, but then they also announced a similar change for Windows 10. And since new features aren’t being added to Windows 10 any more (through its October 2025 end of life), enabling those updates to be installed easily via enablement packages, that makes Windows `10 easier to keep up to date than Windows 11.
Since Microsoft doesn’t indicate (and therefore no one really knows) what constitutes a feature update requiring a full in-place upgrade versus those that can be “flipped on” via an enablement package, Microsoft could have made all the Windows 10 non-feature feature updates require a full in-place upgrade. It might have seemed silly, but it would have put Windows 10 and Windows 11 on equal footing.
My biggest fear here is that IT pros are forgetting how to do in-place upgrades because Windows 10 isn’t making them exercise those skills, so when they do finally upgrade to Windows 11, it will be just as painful as when they started with Windows 10 feature upgrades (so allow plenty of time for that upgrade).
What are the biggest changes in Windows 11? For the initial release, I posted a fairly complete list here and here. But really, the only significant changes were to the UI, with a new Start menu and task bar experience. While I do actually like the new UI, it definitely had some rough edges (and in a few cases, less functionality than what it replaced), but the biggest problem was the extent of the change — IT people aren’t fans of big changes (which is one of the reasons that Windows as a service was initially implemented: so that changes could be incrementally delivered, instead of having one “big” change drop on users).
There were some nice changes related to snapping apps, multiple monitors, virtual desktops, etc., but those aren’t very heavily used. And there were plenty of updates to in-box apps, but most people would rather those just went away. In fact, one of those, the Teams consumer app that is installed by default, is a perfect illustration. For customers already using the work/school version of Teams (don’t ask me why there are still two different apps), this in-box Teams consumer app is terribly confusing to users, so IT has to remove the app — and that’s way harder than it should be. (This goes back to some of the initial Windows 10 pains: including extra “crap” in the OS is one thing, making it hard to remove is entirely necessary.)
A new release should have new features that people want, right?
Granted, Microsoft had some different goals here. The UI changes were to generate buzz (and maybe it didn’t generate the right kind of buzz), but the bigger goal was to get customers to actually use the security-related features that were released in Windows 10. (What features have the least value? Those that don’t get used.)
Lack of an LTSC release?
There have been two Windows 11 releases already, and no long-term servicing release. At this point, you have to wonder if there ever will be one. But is that holding back organizations from moving to Windows 11? Given the small percentages of devices that are actually running an LTSC release of Windows 10, that’s highly unlikely.
It’s a given: Every new Windows release (especially those deployed via an in-place upgrade) is going to break something, whether it’s an app, a driver, or something else. But I haven’t seen or heard anything that would indicate that Windows 11 was any worse than previous feature updates.
There does seem to be a new wave of ISVs indicating that they will “add support for Windows 11 when our customers demand it,” which is definitely a step backwards, but that’s like just a sign of lazy ISVs.
As I said initially, it’s complicated, so there are likely other factors at play. Marketing, customer sentiments (word-of-mouth views), lack of customer focus (no time due to Covid/work-from-home complications), lack of Microsoft focus (Azure rules, Windows doesn’t)? Thoughts?
Categories: Windows 11