Microsoft announces Windows 10 Anniversary Update for summer

[Corky34]

Percentage share doesn't indicate numbers, if there is more device now 12% of the number could be more than 18% at the same time with windows 7.

And that's why it's more correct and honest (IMO) to use a percentage, the "share" part doesn't really come into it.
If you don't use a percentage you're ignoring the fact that your market size has increased, that you should be selling more because you have more people to sell too.

[scaryjim]

... 8 months after launch Windows 10 has a 12-13% share of the market whereas Windows 7 had around 18-19% market share 8 month after release, maybe my figures are wrong though.

It seems to me that ... they're being very deceptive as like you rightly point out they seem to be counting the number of sales, installs, or whatever, but that's pointless or at least being creative with the numbers and hoping people don't notice. ...

Depends on your point of view. if the market has grown significantly (and lets remember there were 6 years between the Win 7 and Win 10 launches) then market share figures aren't necessarily the be all and end all, particularly if that growth in the market can be attributed to the release of the OS: e.g. Win 10 is particularly good for 2-in-1 devices and tablets, which are often companion devices, rather than replacement devices.

MS really aren't trying to be deceptive or misleading about what they consider "the fastest start in Windows history" - they explain it in the very same sentence. If they'd just said "Windows 10 is off to the fastest start in Windows history, outpacing any other version by over 40%", that would be deceptive. Sure, you might disagree with their definition, but at least they give one, clearly, in plain sight.

As I said, my big question is what devices they're counting. I think they're probably ahead of Win 7 whatever spin you give it, but if they're relying on consoles and phones to give the big lead that's definitely "creative"

[Corky34]

As I said, my big question is what devices they're counting. I think they're probably ahead of Win 7 whatever spin you give it, but if they're relying on consoles and phones to give the big lead that's definitely "creative"

From what i can tell, based on nothing more than reading lot of different articles as AFAIK Microsoft haven't defined it, from what i can tell it seems based on all devices that Windows 10 was used on at least once in the last 28 days, so would include tablets, phones, and the console, although I'd be happy if someone had something a little more concrete than my "based on what others have said" reckoning.

EDIT: Also, speaking strictly from a statistical correctness point of view, Microsoft shouldn't really be comparing Windows 10 with previous versions as the way they measure the statistics has changed, IIRC in the past they counted units shipped and now they count active users, yes I'm being pedantic but the misuse of statistics is a personal bugbear for me.

EDIT2: Found an article that goes into some details on what and how they count active users here.

[DanceswithUnix]

So this is a free update. Does that imply in the future there will be not-free updates?

[scaryjim]

EDIT: Also, speaking strictly from a statistical correctness point of view, Microsoft shouldn't really be comparing Windows 10 with previous versions as the way they measure the statistics has changed, IIRC in the past they counted units shipped and now they count active users, yes I'm being pedantic but the misuse of statistics is a personal bugbear for me.

To be fair that's probably just as true for comparing market share statistics across 6 years - the chances of them being compiled in a comparable way are pretty slim. The PC market tends to change quite rapidly.

In fact, I doubt there's any genuinely statistically rigorous way to compare Win 7 and Win 10 sales. For instance, as a counter-point to the "active devices" problem, I could point out that some people bought several Win 7 licenses when the pre-order deals turned up (£30 for a retail copy IIRC) and just left them lying around in case they needed them in the future - they'd all be included in Win 7 sales for the first 8 months, even though they might not have been used until years afterwards.

Anyone got sales figures for XBox One and Windows 10 capable phones? IIRC Windows 10 phone is only just out of Beta, isn't it? I didn't think it was available on many devices yet...

[scaryjim]

So this is a free update. Does that imply in the future there will be not-free updates?

Heh, perhaps Microsoft's definition of "the lifetime of the device" is "until we release an update we want to charge you for"!

[DanceswithUnix]

I could point out that some people bought several Win 7 licenses when the pre-order deals turned up (£30 for a retail copy IIRC) and just left them lying around in case they needed them in the future - they'd all be included in Win 7 sales for the first 8 months, even though they might not have been used until years afterwards.

I still have an unused Windows 8 license from when they did the cheap upgrade license offer

[Corky34]

To be fair that's probably just as true for comparing market share statistics across 6 years - the chances of them being compiled in a comparable way are pretty slim. The PC market tends to change quite rapidly.

Yes they're not perfect but there a lot better than depending on what the companies selling the products publish, AFAIK neither Statcount or NetMarketShare have changed the way they compile the statistics.

In fact, I doubt there's any genuinely statistically rigorous way to compare Win 7 and Win 10 sales. For instance, as a counter-point to the "active devices" problem, I could point out that some people bought several Win 7 licenses when the pre-order deals turned up (£30 for a retail copy IIRC) and just left them lying around in case they needed them in the future - they'd all be included in Win 7 sales for the first 8 months, even though they might not have been used until years afterwards.

Have there even been any Windows 10 sales? I would guess more than 95% of Windows 10 users have taken advantage of the free upgrade.
I do prefer the way Microsoft is now measuring things though as like you rightly pointed out the old way of counting sales probably counted sales that could have sat on a shelf for months or even years afterwards, it's why i prefer to go by the numbers given by the two main bean counters as they've always measured active users, Statcounter measures visitors to a big list of websites, in effect how active each OS is on the internet, while NetMarketShare measures unique visitors to any of the sites they gather stats from so is probably a better reflection individual OSs that are installed.

Anyone got sales figures for XBox One and Windows 10 capable phones? IIRC Windows 10 phone is only just out of Beta, isn't it? I didn't think it was available on many devices yet...

Not sure what numbers you want to go by but from what i gather the XBox One has sold 3 million and shipped 10 million, i guess only the sold would be included the numbers Microsoft are using, IDK the numbers for Windows 10 phones.

[crossy]

So this is a free update. Does that imply in the future there will be not-free updates?

I think that is pretty much a given. Like most others I think a Windows subscription is on the cards (although I'd expect to get basic functions for free).

Didn't I hear something about a possible "Plus" pack a while ago?

[scaryjim]

... Statcounter measures visitors to a big list of websites, in effect how active each OS is on the internet, while NetMarketShare measures unique visitors to any of the sites they gather stats from so is probably a better reflection individual OSs that are installed. ...

Well .... as long as you're aware of all the caveats that go with that approach. Like the OS is self reported, so isn't guaranteed to be right. And the sites involved are self-selecting, so may well be prone to demographic bias. And the type of devices we browse the web on are open to change - Since Windows 7 launched iOS has actually taken 10% of the OS share on StatCounter, so when you look at the figures on there are you looking at just PCs or are you looking across the whole web browsing market?

As it is, StatCounter lets you limit the stats you look at quite effectively. So I've run the numbers for the first 8 months from release, limiting the stats to Desktop only, which will tell us about that key market - the one that both Win 7 and Win 10 address - without the interference from other Win 10 devices (hopefully*).

Win 10 currently - as of end of March 2016 - has 16.5% of the desktop browsing market. It launched on 29th July, so that is 8 full months since launch. Win 7 launched 22nd October 2009**, so if we look at the same time period that takes us to the end of June 2010 (and actually gives Win 7 around a week more from launch), when it's desktop market share was .... 16.1% - so lower than Win 10s after the same time period.

Ultimately, if you want to tell a particular story you can probably find a reliable publicly-available statistic to support it. StatCounter suggests that Win 10 is genuinely outstripping Win 7 not just in absolute installs but in market share too, if you read the stats in a particular way. And I've not had to do anything particularly creative to get therre.

*caveat: I don't know how the various browsers report the OS when used on Win 10 devices. It's certainly possible that some Win 10 devices that are not "desktop" computers (usually defined as x86 desktops and laptops, although I'd be inclined to include x86 tablets as well) are currently counted in the desktop browser stats. As I said, that's one of the problems of using web browsing data for market share statistics....

** This was the official release to the public, although it went out to MSDN and Technet subscribers in August that year, and actually had 2% of the browsing market share before the official launch - which means that from launch to Junr 2010 it only gained ~ 14%, making it quite a bit slower than Win 10...

[Corky34]

Well .... as long as you're aware of all the caveats that go with that approach. Like the OS is self reported, so isn't guaranteed to be right. And the sites involved are self-selecting, so may well be prone to demographic bias. And the type of devices we browse the web on are open to change - Since Windows 7 launched iOS has actually taken 10% of the OS share on StatCounter, so when you look at the figures on there are you looking at just PCs or are you looking across the whole web browsing market?

Apologises I didn't mean to say or imply that either of the bean counters are an exact or precise reflection on the actual numbers of a particular OS that's installed, if you want that for Windows 10 then Microsoft have published the figures, what they're good for (IMO) is comparing what was once Microsoft's suspicious claims of how well a particular OS was selling or being adopted (ignoring any suspicious claims from other companies).

In the past, like you said earlier, Microsoft counted units shipped even if a unit was sitting on a shelf, like when they claimed 100 million Windows 8 licenses had been sold 6 months after launch, previously, and without using the bean counters, there was no way to verify their claims, compare with previous releases, or how many were actually being used and not gathering dust.

Like i said the new way Microsoft is using to report active users is much better as its a more accurate reflection (IMO) of actual installs/users, my only gripe is that their using numbers from two different methods of collecting statistics so when they claim "Windows 10 is off to the fastest start in Windows history" and "outpacing Windows 7 by 145%" they really shouldn't or can't make claims like that when the method or recording the statistic has changed so drastically.

It why personally I'd trust the bean counters more than any claim that Microsoft makes as their methods haven't changed so your comparing like for like.

As it is, StatCounter lets you limit the stats you look at quite effectively.....
Win 10 currently - as of end of March 2016 - has 16.5% ...... Win 7 launched 22nd October 200 it's desktop market share was .... 16.1% - so lower than Win 10s after the same time period.

Yea the March figures were only updated today (1st of the month) and it looks like Windows 10 put on another 1.5% on Statcounter, historically it seems to have pegged Windows 10 slightly higher than NetMarketShare (14.15% for March), maybe that's to do with the different ways they record their stats (page views versus unique visitors) and Windows 10 having a more talkative nature.

Ultimately, if you want to tell a particular story you can probably find a reliable publicly-available statistic to support it.

Indeed, it's probably why it gets my goat so much when companies and politicians use statistics to support their narrative, and why we have things like the UK Statistics Authority, and the European Statistics Code of Practice.

[WizFiz]

The fastest due to it being "free" as others said. Then you got the rest who like to try something new, or were sucked into thinking DX12 Applications were coming anytime this century.

Should be interesting to see what bugs and performance issues will be introduced with this update.

[shaithis]

The fastest due to it being "free" as others said. Then you got the rest who like to try something new, or were sucked into thinking DX12 Applications were coming anytime this century.

Should be interesting to see what bugs and performance issues will be introduced with this update.

In the past, forums like these would be full of help posts whenever a new version of Windows released. Wherever I have worked, there have always been a slew of users screaming for the new Windows every time it released......

....and this time, the upgrade was FREE. I didn't think there would be anything other then a HUGE uptake.

As for DX12, the games are appearing faster then I ever expected. Quality aside, they are releasing fast. GoW:UE, AotS, Hitman, RotTR and Quantum Break to name the big hitters off the top of my head.

[madmaniakid]

The only reason it's outselling all other versions is because you cannot buy any other version when 10 was released, whereas with the other versions you could still buy the previous ones.

[jigger]

I've had one machine update to Windows10 on it's own!

[jigger]

Is the HoloLens Windows 10 only? I would like to get one of these at some point.