Microsoft should have Windows 10 on more than 410 million personal computers within 18 months of the new OS's release, an analysis of user share data and upgrade tempos shows.
More than three-fourths of all PCs running Windows 8 or 8.1 will migrate to Windows 10 in the first 18 months of the latter's lifecycle. A smaller percentage of Windows 7 PCs will add hundreds of millions more to the tally, even though most business and government devices will not shift to Windows 10 until 2017 or later.
The calculations are based on the performance of Windows 8.1, the free upgrade Microsoft shipped in October 2013 as the follow-up to the original Windows 8 of the year before.
According to analytics vendor Net Applications, which measures operating system user share by tracking unique visitors to its customers' websites, 76% of PCs running Windows 8.1 or Windows 8 ran the former in April, an increase of about 11 percentage points from six months before and 26 points compared to 12 months prior.
Windows 8.1's adoption has been unprecedented for a Microsoft OS -- not a shock since it was also the company's first free upgrade available to all users of Windows 8.
Microsoft is counting on the same trend for Windows 10: Two weeks ago, the company publicly set a goal of having the new OS running on one billion devices within two to three years. To meet that target, analysts have calculated, Windows 10 will not only have to capture virtually the entire consumer new PC market during the 24-to-36-month stretch and power a larger number of smartphones, but also convince a significant chunk of Windows 7 users to upgrade to 10, either by purchasing a new device or taking advantage of Microsoft's free upgrade offer.
With Windows 8.1 as a guide -- and the fact that Windows 10 is also free -- it's a good bet that 76% of the former at the time the latter launches will be upgraded within 18 months. In other words, if Microsoft launches Windows 10 in July, it should be able to get three-fourths of all Windows 8 and 8.1 PCs (as of that month) onto the new OS by the end of 2016.
Assuming 1.52 billion Windows PCs worldwide, that translates into 198 million machines (17.1% share for Windows 8/8.1 at the end of June x 76% = 13% x 1.52 billion).
Figuring Windows 7's migration is more problematic. Research firm IDC says that approximately 55% of all PCs are in the hands of consumers, 45% in commercial organizations. If that ratio applies to Windows 7 -- which will account for 67% of all Windows PCs by June's end -- that means consumers' systems will represent about 37% of all Windows 7-powered machines.
Commercial PCs would be discarded from the projections: Even though they will have to eventually upgrade to Windows 10 -- the enterprise standard, Windows 7, exits support in January 2020 -- they won't begin seriously upgrading until 2017 or 2018, according to analysts. That's after the 18-month forecast timeline.
It would be foolish to assume Windows 7 PCs, even those owned by consumers, would upgrade to Windows 10 at the same rate as those running Windows 8/8.1. Windows 7 runs on about four times as many systems as 8/8.1, and moving that massive number will be difficult and time consuming.
If, say, consumer Windows 7 PCs upgraded to Windows 10 at just half the rate of Windows 8/8.1, and the upgrades begin in July when Windows 7 should account for 67% of all Windows systems, about 14% of all Windows machines would migrate to the new operating system within 18 months (67% x 55% for consumers' part of the total, x 38% upgrade rate, or half the 76% of Windows 8.1).
That 14% would represent 212 million PCs.
Together, the Windows 8/8.1 and Windows 7 upgrades would total 410.8 million within a year and a half, and at the end of the period account for 27% of all Windows PCs.
The projection may be on the conservative side: Although the 27% would be a migration speed record for Microsoft, beating the 25% uptake of Windows 7 in its first 18 months, the slight advantage to Windows 10 may be underestimating the shift, since Windows 7 was not a free upgrade, and Microsoft was not subsidizing ultra-cheap notebooks, as it is now.
But even sticking to the 410 million gives Microsoft a good jump on its 1 billion devices goal, assuming that the length of Redmond's forecast is three years, not two. The three-year timespan also will have the added advantage of including major rounds of enterprise migration, since by the end of the stretch companies will have only a year and a half before Windows 7 gets put out to pasture.
"Enterprises will use the fourth quarter of 2015 as a learning experience," said Stephen Kleynhans, an analyst with Gartner, in an earlier interview about upgrading to Windows 10. "They'll watch what happens on the street with upgrades. They'll see 2016 as an opportunity to roll up their sleeves and start. Then if nothing horrible happens, in 2017 they'll start a slow ramp-up to the product, a bit of a slow ramp at first, then faster. The real deadline, of course, is 2020."