Microsoft's big Windows 10 goal: one billion or bust

Earlier this month, Microsoft executives laid out a lofty goal: one billion devices running Windows 10 within two to three years. That sounds like a big number, but how tough will it really be to hit that target?
Written by Ed Bott, Senior Contributing Editor

For Windows 10, Microsoft is aiming for a 10-digit installed base.

That's a 1 followed by 9 zeroes: 1,000,000,000. One billion devices running Windows 10.

At the Build developers' conference in San Francisco earlier this month, Microsoft's Terry Myerson declared that the company's goal was to reach the billion-device milestone "within two to three years after launch."

Now a billion might not be what it used to be, but it's still a big number. How realistic is it?

If you do the math, it's not hard to get to that number and even beyond.

That math is, unfortunately, fuzzy, because we don't really know how many PCs and tablets and phones are out there to a fine level of precision, nor what the owners of those devices have done with them since they were purchased.

But there are enough reliable sales figures out there, some in official reports from companies, others supplied by trustworthy market research firms, to make some educated guesses. Even with wide-ranging assumptions and allowing very broad margins of error, there are plausible paths to a billion.

With those caveats, let's survey the Windows 10 landscape as it might look in 2018.

New Windows 10 PCs and tablets

400-800 million

Every quarter, Gartner and IDC report their estimates of worldwide PC sales for the previous three months. They also publish projections of that the market will look like a few years into the future.

The most recent numbers show steady declines in sales of traditional desktop and laptop PCs, with steady growth in the (currently much smaller) segment of what Gartner calls "premium ultramobiles" like Apple's latest MacBook and Microsoft's Surface Pro series.

Consensus estimates of Windows PC and tablet sales for the next few years are flat, at roughly 300 million annually.

Not all them will be running Windows 10, of course. Microsoft has yet to announce an end to sales of Windows 7 PCs, and there are plenty of enterprises that will exercise downgrade rights to maintain their Windows 7 deployments.

Even if only half of those 300 million new PCs each year are running Windows 10, that's nearly half a billion after three years. Given Microsoft's willingness to deal on the cost of Windows 10 and its ability to end sales of Windows 7 PCs, the actual percentage of new devices running the new OS will probably be much higher.

New and upgraded Windows phones

200-500 million

Windows Phone is a dud, right?

Yes, if you measure success by market share. Microsoft's mobile OS, Windows Phone 8.1, has a paltry share in the U.S. (well under 5 percent regardless of which market research firm you ask) and is doing a only slightly better elsewhere, well under 10 percent worldwide.

But the market for smartphones is so big--roughly 1.8 billion per year--that even a small share represents a lot of units. In the most recent quarter, Microsoft sold 8.6 million phones worldwide. That should be a worst-case quarter, and even a weak response to Windows 10 coupled with new flagship devices would put total worldwide sales at between 50 million and 150 million Windows 10 phones per year.

Add in another 50 million or so current Windows 8.1 phones that are eligible for upgrades and you're up to a couple hundred million devices in a platform that is a distant third to iOS and Android.

Consumer and small business upgrades

400 million+

Back in 2012, a few months before Windows 8 shipped, Microsoft announced that it had sold 600 million licenses of Windows 7, and given the tepid reaction to Windows 8 it's likely that they've sold at least another 200 million Windows 7 licenses since then. Some of those PCs have been retired and some have been upgraded, but it's still likely that the current Windows 7 installed base is at least 600 million.

Microsoft hasn't announced sales figures for Windows 8 and 8.1 licenses in a long time. More than a year ago, that number was 200 million. Let's be pessimists and assume that that figure has gone up only 50 million in the past year.

I estimate that 90 percent of Windows 8.1 PCs will take advantage of the free upgrade to Windows 10. That's at least 225 million PCs right there, most of them very grateful to have a Start menu back.

So what percentage of Windows 7 PCs will upgrade? That's a tougher sell for Microsoft, which will probably be thrilled if half of all Windows 7 PCs move to Windows 10. But let's assume, conservatively, that only 1 in 4 Windows 7 PCs are upgraded. That's still at least 150 million PCs.

Upgrading 400 million PCs, most of them in the first year, would be a tremendous achievement, the largest single software upgrade in history. And yet that's at the core of Microsoft's plan.

Enterprise upgrades


This one's a real wild card. Under normal circumstances, only about 10 percent of enterprise customers would deploy a new version of Windows within two years of its launch. It's possible, even likely, that this time will be different.

First of all, Windows 10 is really just a straightforward update to Windows 8.1, which in turn was a straightforward update to Windows 8. Second, there are some compelling enterprise features in Windows 10, including Device Guard and the ability to package Win32 desktop apps and deliver them through the Windows Store.

But for the sake of this estimate, let's not even include a number here.

After all that, even with relatively pessimistic assumptions, there are roughly a billion Windows 10 devices in use in mid-2018. If Microsoft can keep Windows 10 Mobile device sales at 10 percent of the worldwide market or more and help its OEM partners sell more Windows 10 PCs and tablets, that total increases by hundreds of millions more.

And I didn't even mention pirated copies of Windows.

Execution is everything, of course, and putting that 10-digit number out there as a goal is actually defining the minimum acceptable standard of success. Let's check back in two years and see how it all worked out.

Editorial standards