Microsoft is declaring the first month of Windows 10 a success. But what conclusions can we really draw about the reception of this critical update?
Technology companies can be rather frugal with concrete information from time to time. This is understandable, because hard numbers sometimes misrepresent a good situation as bad. However, they can also represent a bad one as exactly what it is, unfortunately.
As a general rule, if a business is giving chapter and verse on how many of something it’s sold, it’s usually good news, and if it’s being very vague or just not saying anything, then it’s invariably a less wonderful story to tell.
That’s precisely the context in which to place the recent drum banging by Microsoft about Windows 10, because it comes on the back of an effective news blackout on Windows 8 sales that we can all understand it would like to move on from.
For the first month, it’s been a solid start, because neither Windows 7 nor 8 could claim 75 million installations over this period, and that’s how many Windows 10 installations Microsoft claims to have been bagged.
But conversely, Windows 10 is free to owners of the previous two operating systems, so that’s clearly a factor in convincing people to make the move. And Windows 8.x wasn’t much liked by the general populous, making for yet another rogue component.
Is convincing so many people to adopt their new OS proof that Windows 10 is a winner for Microsoft? Or is it all a bit more complicated than the obvious and immediate conclusion to draw?
To get a better grip on this question, we need to go back to Windows 7 and 8, and build some sort of model for how Windows transitions happen and compare that with what we know about Windows 10 so far.
Windows 7
According to the best information available about Windows 7 sales, Microsoft sold 60 million licences over the first two months after launch, and it then went on selling roughly 20 million copies a month for the next three years or so.
As operating systems go, Windows 7 was massively successful for Microsoft, coming as it did after the woeful Vista episode.
Within the first six months, 10% of PCs in the world were running Windows 7, and eight months after that 20%. In fact, the percentage of computers running Windows 7 has grown almost continually since, even after Windows 8 launched.
However, what’s different about the 60 million copies number here is that this wasn’t installations; it was licences shipped, and that is something quite different.
This is a trick phone makers often use to make their sales seem bigger, where they count every phone that has shipped to a phone reseller as sold. But they’re not, and it’s likely that after the initial delivery there won’t be any more shipments until that stock is sold.
Windows 7 was still in an era where retail disc versions were shipped to shops where they sat on a shelf, waiting for happy punters to come along with hard cash.
What some have also forgotten is that because XP was so popular and Vista much less so, Microsoft decided, not unlike Windows 10, to encourage its customers with some early discounts.
You could get a cheap Windows 7 Home upgrade, and it also did a cheap ‘Family’ triple pack for about the cost of a single standard licence.
No, it wasn’t ‘free’, but it wasn’t full price either, and like Windows 10, it’s predecessor wasn’t well received.
The sales of Windows 7 are best described as a slow burn, where people didn’t charge en masse to it, but when the opportunities came along, they often moved. That’s especially true of Vista installs, because those numbers dwindled to almost nothing in relatively short order. Whereas the number of Windows XP users is still many tens of millions, the number of Vista users left now is miniscule.
After Windows 7, many in Microsoft must have thought its flagship product was an untouchable brand, but its next upgrade proved exactly the opposite.
The sales of Windows 8 were so bad that Microsoft made great efforts to obscure the true numbers, mostly out of utter embarrassment.
Windows 8
Based on Windows 7 numbers, Windows 8 needed to ship more than 30 million in its first month to show its predecessor a clean pair of heals. And, amazingly, it did that.
According to the numbers Microsoft released, 40 million licences shipped in the first calendar month, but again this is licences shipped, not sold or installed. Many that did sell went to businesses with no intention whatsoever of installing Windows 8.
Pointedly, Microsoft gave no breakdown of how many were installed or shipped pre-installed on computers. Two things gave away the true situation that only those inside Microsoft knew, one of which was the last official numbers on Windows 8 that it issued.
They next came in June 2013, when Microsoft declared that the 100 million mark had been reached. That was the last report on the total sales Microsoft would ever make about Windows 8 sales to this day.
It doesn’t take the maths skills of Stephen Hawkins to work out that if you sold 40 million in month one and then eight months later you’d only sold another 60 million, then something had gone very badly wrong indeed.
What had happened was that after the initial launch, sales plummeted like Microsoft was selling a contagion, and they never recovered in any meaningful way even when Windows 8.1 was introduced.
It’s worth remembering that during this entire fiasco, Microsoft was making lots of positive noises about how happy it was with Windows 8 sales, when it knew it was facing a total meltdown.
Based on statistical data gathered by analysts beyond Microsoft’s control, it’s been estimated that whereas Windows 7 took six months to reach 10% of the PC market, it took Windows 8 a whole 15 months to hit that level. At that point the sales curve went almost flat, and it never really got much above that, peaking at about 12%. At the rate it was gaining on other versions of Windows, it was estimated that it might take anything up to 150 years to become the most popular release.
By comparison, even the generally reviled Vista managed a peak of 23.6% of the Windows market, before Windows 7 arrived.
For proper balance, I should mention that there were more computers in the world in the Windows 8 era than the Windows 7 one, so reaching the percentages took more sales, but that doesn’t really make up the difference. Windows 7 peaked at more than 50% of all systems at one point, and Windows 8 never threatened to overtake Windows XP or Windows 7.
But there was another clue about when Microsoft realised Windows 8 wouldn’t fly, and it was the departure of Windows 8 senior software architect Steven Sinofsky just three months after Windows 8 launched. His exit, which according to many was a mutually agreed move, can only been because of Windows 8 sales and his part in the design so many people disliked, because whatever personality traits Sinofsky had, he’d been at Microsoft since 1989 and led the Windows team since 2009, so they all knew who he was and how he operated.
That he left on 31st December 2012, to take $14M in stock for his one-year incomplete contract, and various internal sources identified him as the reason that Windows 8 failed so completely.
That it happened then demonstrates that even at that stage Microsoft knew how these runes would land in the coming months, and with Steve Ballmer looking to hand over to someone else, he was the first person the new Microsoft wouldn’t be requiring. He handed over to Julie Larson-Green and Tami Reller, so perhaps it would be useful to keep an eye on their career paths around Christmas time.
In the end, mostly because it was an elephant in the room larger than the state of Texas, the company finally admitted that Windows 8 hadn’t been the rip-roaring success it had hoped for.
However, it took until October 2014, nearly two years after it parted company with Steven Sinofsky, for new CEO Satya Nadella to say, “Let’s face it, we got some things wrong in Windows 8.”
So based on both these prior models, is Windows 10 going the way of version 7 or 8?
Windows 10
I find it highly refreshing that Microsoft chose to release numbers on activated installations, because that’s surely much nearer the real picture than some dressed-up quirk of accounting being passed around as gilt-edged success.
And, as activated installations go, 75 million in the first month is a pretty healthy situation. However, as Windows 8 well demonstrated, the first month isn’t really representative of success, because of the numerous special circumstances that surround a new launch.
In this instance, being ‘free’ is one, although if people hate Windows 10 and revert their licences to Windows 7 (or 8, disturbingly), then that could really mess with the statistics from here on.
What Microsoft will never know is how many of those installations would have happened if Windows 10 cost money at this stage. Information it might have, but isn’t sharing is the breakdowns of in-place upgrades. If the vast majority of these are Windows 8 users desperate to get away from that release, then that’s a lot less encouraging than if they’re Windows 7 users ready for a fresh OS.
Based on my own experience, in that the only machine I’ve upgraded so far is a Windows 8.x one, I suspect that many of these installations are erasing Windows 8 at this point. Only Microsoft knows the real situation, and it hasn’t fired anyone senior yet.
With this being the first version of Windows that’s ‘free’ to a large chunk of existing users, Microsoft needs to hold its nerve and watch what trends develop over this special year – because even if Windows 10 only cannibalises all the Windows 8 licences over that time, it will have a good chunk of the overall market, and it will have at least bettered Windows 8 for how rapidly it got that many together.
To also put this into perspective with Apple, which it’s generally accepted has a user base of around 100 million worldwide, by the time you read this, there will be more Windows 10 systems in use than those running Mac OS X globally.
Having just watched the World Athletics Championship, the Windows 10 situation reminds me much of those who see GB athletes win medals in Beijing and then tick those same boxes for success in the Rio Olympics.
Those battles have yet to be won, and in many respects Windows 10 has yet to even start on ticking the box that puts Windows back in the position of grandeur it once occupied.
Given how the world of computing has radically changed in the past decade, maybe Microsoft’s expectations are rightly more modest.
Therefore, it might be cheered by the response to Windows 10 so far, but also mindful that it’s in it for the long haul and not to celebrate before it’s really confident about what it’s seeing from both a bought licence and user activity statistics.
We’re a long way from erasing Windows 8 from the minds of many, though at this time Windows 10 appears to have taken a few confident first steps.
Only if Microsoft can entirely unify the Windows world around this OS can it ever be classed as a true success, and given the generally positive vibe people still have about Windows 7, that might be a tall order.
I’ve no doubt it can easily cannibalise the Windows 8 user base, but there is still a big question about going beyond that objective and how the year of free upgrades is likely to change the perceived value of this product in the longer term.
What it can’t address, however well it is ultimately received, is that since iOS and Android began their respective rises, the concept of a full desktop OS has taken a bashing. With so many people doing many of their daily computing tasks with just a phone or a tablet, the necessity of a desktop OS has been severely undermined.
Windows 10 might have had the dream start that Microsoft would have hoped for, but that doesn’t imply that it can undo the change that’s occurred in the past five years to the fabric of computing.
Windows Sales Numbers (With Source)
• Windows 1.0 sales from its November 1985 launch to April 1987: 500,000 (Computerworld).
• Windows sales in 1988 (Windows 2.0 shipped on December 9, 1987): one million (InfoWorld).
• Windows sales, all versions, 1985 to January 1990: less than two million (InfoWorld).
• Windows 3.0 sales, first year: four million (InfoWorld).
• Windows 3.1 sales, first three months or so: three million (InfoWorld).
• Windows 95 sales, first year: 40 million (Network World).
• Windows 98 sales, first four days: 530,000 boxed copies through retail channels (New York Times).
• Windows 2000 sales, less than a month after launch: one million (Microsoft).
• Windows ME sales, first three days: 200,000 boxed copies through US retail channels (Network World).
• Windows XP sales, first three days: 300,000 boxed copies through US retail channels (Network World).
• Windows XP sales, just over two months after launch: 17 million (Microsoft).
• Windows XP sales after five years, 400 million (IDG) .
• Windows Vista sales, one month after launch: 20 million (Microsoft).
• Windows Vista sales, 180 million after 18 months (Microsoft).
• Windows 7 sales, first six months: 100 million (Engadget).
• Windows 7 sales from October 2009 launch to June 2010: 150 million (Neowin).
• Windows 7 sales in less than two years: 450 million (TechCrunch).
• Windows 8 sales in a little over two months: 60 million (ZDnet).
• Windows 8 sales, first six months: 100 million (Microsoft).
• Windows 10 installs, first month: 75 million (Microsoft).
Stacking The Deck
Anyone who’s worked in a big company knows that those with the fiscal controls often present a different profit and loss picture than the one that accurately represents the state of play.
One of the classic tricks is to defer taking profit at one point in the year, to then pull it like a rabbit out of a hat later on. The basic logic of this is if news is bad, then make it as bad as it can be, and then at the next performance announcement, confuse stockholders by making the numbers correspondingly better than expected.
So how does this relate to Windows and Microsoft? Using Windows 8 as a classic example, here is how it all works.
In 2012, prior to the launch of Windows 8, Microsoft deferred $540m of revenue from Q2 of that year, and then another $1.36bn in Q3 from the Windows division. The excuse given for this was that it intended to give a cheap Windows 8 upgrade to anyone who bought Windows 7 close to the launch, and that was to cover that cost.
This assumes they’ll all upgrade, which Microsoft knew wouldn’t happen, and it also only took the revenue when they did the upgrade or on 28th February 2013.
In this way, Microsoft had $2bn in the back pocket which by March 2013 would come back on to the profit and loss account for the Windows division even if it didn’t sell a single copy of Windows 8.
In addition to making the Q1 2013 sales of Windows 8 look impossibly good, this also had the added benefit that it made the actual number of copies sold almost impossible to calculate for those trying, like most tech analysts would be.
However, there are limits to how much you can play these games, and those that push the boundaries can incur the wrath of the US Securities and Exchange Commission, as well as their own stockholders for misrepresenting their profitability.
Microsoft chose not to go there, admitting that Windows 8 wasn’t selling as many copies as it had hoped, but it actually avoided saying just how badly it did.
Additionally, all new Windows releases generally get a leg up, because from shortly after the point of release, all new PCs come with the new OS. However, this point went rather wrong with Windows 8, when end users started demanding Windows 7 or, in the case of business users, exercised downgrading rights.
It’s generally accepted that because the usage figures for PCs browsing the internet didn’t match those licences that Microsoft claimed to have sold, a great many PCs got downgraded.
This why it’s so important that Microsoft is now stating installed and activated numbers, because otherwise the sold licence figures are open to whatever interpretation you wish to apply to them.
What’s also different is that Microsoft has changed the way that it accounts for Windows revenue, choosing to take the amount it receives at point of sale and spread it over the next three years. This, and the relatively low numbers of actual Windows 10 sales in the next 12 months (rather than free upgrades) will probably make the numbers for this division even more confusing to unravel.