Windows 7 is the weekday workhorse, but Windows 10 and 8.1 are what people choose to use
Recently, my fellow ZDNet blogger JackSchofield looked at what the StatCounter figures say about the continuing decline of Windows XP and whether Windows 10 has passed XP yet. Given that XP dates back to 2001 and wasn't designed for the security threats of the modern iinternet, the faster it declines, the better. But if you look at the daily statistics as well as the monthly ones, some other interesting points emerge.
In particular you can clearly see the weekly usage patterns -- and for different operating systems, you can see a clear peak or drop at the weekends. Note: I'm discounting personal use in the evenings because there's no way of seeing it in the figures, and these charts run from July 3rd, a Friday, to November 3rd, a Tuesday, to make the weekly pattern as clear as possible.
Windows 7 is a weekday workhorse; people use it at work and the dip at the weekends is clear -- and it's been getting steadily more visible since Windows 10 launched at the end of July. The same is true of Windows XP, at least in the worldwide figures (as we'll see, there are some variations worldwide). These are operating systems people still use at the weekend, but there's a clear percentage of users who only use Windows 7 or Windows XP during the week, likely at work.
OS X shows the same weekend dip, at least worldwide -- again, that varies by region, as does whether it's ahead of Windows 10 (North America), behind (Europe) or neck and neck (worldwide but Windows 10 has been pulling ahead in the last week).
But Windows 8.1 -- which is still more popular than Windows 10 -- Windows 10 and even Windows 8 (to a much lesser degree) peak at the weekends. These are operating systems people are choosing to use -- or using on the computers they choose to use at the weekend.
In the US, the peaks and dips are similar, even though the comparative share is different. Windows 10 and Windows 8.1 peak at the weekends. Windows 7 usage dips at the weekends, far more sharply than in the worldwide figures. Windows XP sometimes peaks at the weekend, sometimes dips at the weekend and generally wobbles around at a low enough percentage to suggest that the sample may not be statistically representative. OS X and Windows 8.1 usage duel in the US, with the switch to Windows 10 leaving OS X usage above 8.1 by September. And OS X again shows weekend peaks, although not as cleanly.
Windows 8.1 is only just above Windows 10 in Europe as of the beginning of November. And Windows 10 passes OS X usage in Europe sometime after the middle of September; the weekly bumps and dips make it unclear at first but by the middle of October, Windows 10 usage is above OS X usage in Europe even. In the week, when OS X usage is highest and Windows 10 usage lowest.
The weekly patterns are less pronounced in Africa but still visible where Windows XP and (sometimes) Windows 8.1 are the weekend choice and Windows 7 usage peaks in the week.
South America has very strong weekly patterns; you can watch Windows 7 lose share at the weekends as Windows 10 arrives, with increasing weekend peaks. By the end of October, these exactly mirror the drop in Windows XP usage at the weekends. Windows 8 peaks up at the weekends in South America; OS X dips down at the weekends (it's about to pass Windows 8 usage because of the switch to Windows 10 but it remains below even Windows XP).
Asia doesn't show the weekly behaviour as clearly for Windows 7, although the weekend dips are beginning to be visible. Here, Windows 8.1 and Windows XP are duelling, with 8 peaking at weekends as XP drops at the weekends. Windows 10 growth is slower but again, it goes up at the weekends, when Mac OS X usage goes down.
Australia and New Zealand -- where OS X share is high but again only dropped below Windows 8.1 because of the switch to Windows 10 -- show the weekend Windows 7 dip very clearly. Here, OS X, Windows 8, and Windows 10 are all strong weekend choices, but Windows 10 has started to match Windows 8.1 usage at the weekend, though not yet in the week. XP shows a slight weekend dip and Windows 8 a slight weekend peak down under.
I haven't included the daily chart for Antarctica, because there are so few people at the bases at this time of the year that the statistics seem to depend on who has come in on the supply plane each week and what computer they happen to turn on, but the percentages are interesting. It's the only place where Linux makes it into the top seven operating systems (lots of scientific computing going on), but it's in third place behind Windows 7 and Windows 10.
You can make your own charts at StatCounter. Remember that the weekly and monthly charts smooth out the variation between the week and weekend, so make sure your charts begin and end at the same time of the week for the most accurate comparison. And remember that StatCounter measures the amount of web surfing people do rather than just the number of users, although its trends are broadly similar to Netmarketshare, which tries to measure the number of unique PCs.
I do disagree with Jack slightly on his expectations about continuing upgrades to Windows 10. He thinks free upgrades will taper off because those who want the free upgrade will have jumped already. The comments on my Uservoice request to replace placeholders in OneDrive suggest I'm not the only person keeping Windows 8.1 around because the experience on a tablet is still somewhat better (Windows 10 has certainly improved this and continues to do so) and because of placeholders.
Note: the Uservoice site for Windows 10 features is being retired in favour of the Windows 10 feedback app -- although the developer-focused sites for the Windows 10 platform and Edge will carry on -- so the comments have already been removed and that link may stop working after mid-November.
It's been a year since placeholders were removed from Windows 10 (back during the technical preview) and while new OneDrive head Jeff Teper mentioned to me on Twitter that the request has certainly been heard and Microsoft knows there is "more to do", there hasn't been an official statement on the plan to deliver something else to address the same scenario since January 2015.
Microsoft is clearly reconsidering its plans for what the OneDrive cloud service is intended to be, and when it either finds a solution for placeholders or explains what else users should do, I expect to see another bump in Windows 10 upgrades. And if that hasn't happened before the end of the free year in July 2016, I'd bet on a lot of us Windows 8.1 holdouts jumping to 10 anyway, to get the free upgrade. Thanks to Windows as a service, Windows 10 will have improved by then anyway.
Now read this
- Windows Mobile: Can Microsoft get us excited again?
- Open source: Free as in speech, beer - or puppy?
- Diary of a Microsoft Band user: Here's what I've found out so far
- How three years of Yammer has changed Microsoft for the better
- What 'client' means on the new Windows Server
- Pressed for Time, book review: Don't blame the technology
- How Clutter saved me from inbox 4,000