A month in, Microsoft has made it 1/13th of the way to its two year goal -- let's look at what the future holds

A month ago to the day, Microsoft Corp. (MSFT) launched Windows 10 to the masses.  At the helm was Yusuf Mehdi (@yusuf_i_mehdi).  In June Mehdi, a senior vice president (SVP) at Microsoft, was promoted to head the Windows and Devices Group amid a broad shakeup of Microsoft's executive ranks.

I. A Record Setting Pace

A month later, the new Windows chief has happy news to report -- as of Wednesday, Windows 10 has seen more than 75 million installs.  Mehdi announced this milestone as part of a series of post to his Twitter Inc. (TWTR) account: Note that it took Windows 7 nearly three months to achieve 60 million license sales.  So Windows 10 is being adopted 3.75 faster than Windows 7, Microsoft's fastest-adopted OS to date!

While this is an important milestone, Microsoft has its sites set on a more distant target and a far bigger number -- a billion users.  CEO Satya Nadella in April said that Microsoft is gunning to gain a billion total users of Windows 10 within the operating system's first three years on the market (so by the end of July 2018, roughly).

I took the liberty of graphing Microsoft's growth including the following events: Windows 10

As you can see above, Microsoft's early install pace has been very aggressive, thanks in part obviously to the fact that Windows 10 is free to owners of a Windows 7, 8, or 8.1 device.  If it were hypothetically able to sustain this pace of growth (70 million new users added per month), it would reach its target nearly two years early, sometime in next October.

Practically speaking that seems improbable for Microsoft to sustain that kind of growth in the PC space, but there's some chance that it may be able to achieve such a feat in terms of total installs.

Let's examine what the future holds.

II. Xbox One

First, remember Microsoft is going to get a significant boost in Windows 10 use, when it rolls out an upgrade for Xbox One users.  Currently the Xbox One has sold somewhere in the neighborhood of 14-17 million units, in the 22 months since launch.  By mid-2018 I expect that somewhere around 40-60 million units to have been sold.

Xbox One growth

Xbox One

Xbox One growth
Xbox One -- sales growth [Image Source: ArsTechnica]

So likely at least 5 percent of the target will be met simply by upgrading Xbox One users.  As consoles have near 100 percent firmware upgrade rates, I don't think this will be much of a challenge.

III. Windows 10 IoT Core

Then there's Windows 10 IoT Core -- a slimmed down series of Windows 10 customizable interface stack for use on devices like the Raspberry Pi.  This set of lightweight libraries is already available.  The Raspberry Pi Foundation announced in February of this year that it had shipped 5 million of its tiny computers to tinkerers in its first three years of sales.

Raspberry Pi Model 2
The Raspberry Pi Model 2 B

Along with similar devices, such as the Intel Corp. (INTC) "Edison" IoT platform and the MinnowBoard Max, I'd conservatively estimate that by mid-2018 there will be at least 15-20 million of this class of nano-PCs in the wild.  I'd further wager that at least 10 million of the devices will have been used in conjunction with Windows 10 IoT .  If I'm right that'd add another 1 percent towards Microsoft's three-year goal.

Microsoft's augmented reality device -- the HoloLens -- another small contributor could see sales as high as 2-5 million units between mid-2015 and mid-2018.

IV. Windows 10 for Smartphones and the HoloLens

Turning to the smartphone space, Microsoft finished 2014 with an estimated 35 million units in the wild.  While usage grew nearly ten-fold in 2013, according to market research firm AdDuplex...

WP8 and WP7
[Image Source: AdDuplex] 2014 the lack of new high-end product and a coherent sales message hurt the Lumia line's growth.  In the April-June period (FY Q3 2015), Microsoft sold 8.6 million Lumia units [source].  While that represents an 18 percent increase over the same quarter in 2014, it's a far cry from the meteoric growth the platform was seeing in 2014.

There's a fair amount of uncertainty surrounding the platform, particularly with Microsoft committing to yet another round layoffs of Lumia engineers in July.  Among the factors that are unclear include:
  • What will the adoption of Windows 10 on Android replacement devices be?
  • Will third parties embrace Windows 10 in the smartphone space, given its free license for commercial use in small to mid-size mobile form factors?
  • What will the fate of the Lumia unit (and its sales) be given the cuts?
In three years we'll certainly know the answers to those questions.  But for now, I would argue it's all but impossible to guess.  Even if you knew Microsoft's entire internal gameplan, the picture is still very well might shift so far in the next three years that it would still be impossible to know the ultimate answers to those questions a priori.  Suffice it to say, the Windows Phone line is at a time of tremendous promise, but remendous chaos as well.

Still Windows Phone should comprise a significant percentage of Windows 10 installs, in part because the platform has a good track record when it comes to system updates.  By the end of 2013, for example, more than 70 percent of users had moved on to Windows Phone 8.

WP8 vs WP7

So by mid-2018, we can probably estimate that Windows 10 will be on at least 90 percent of Windows-powered smartphones.  And based on history, in spite of all the chaos, I'd argue you could estimate conservatively that there will be around 45-50 million Windows 10 powered smartphones by that time, based on market growth, even if the platform only manages to hold steady amidst the chaos.  If I'm right, smartphones will represent at least 4 percent of the billion device mark Microsoft is gunning for.

V. The Tablet

That brings us to the final categories -- PCs and tablets.

Gartner, Inc. (IT), a veteran market research firm, estimates that around 318 million and 322 million PCs will be sold for the year of 2016 and 2017, respectively.  For the same period, Gartner sees sales of budget "ultramobile" class devices (which includes most tablets and clamshell/netbook type designs) reaching 258 and 276 million.

The International Data Corp., Gartner's chief rival, offers a slightly drearier prediction based on the recent currency-driven economic downturn.  Note that it categorizes devices slightly differently, counting budget clamshells and hybrids as PCs.  It originally predicted sales of 374 million and 382 million desktop and "portable" class PCs in 2016 and 2017, alongside 319 million and 352 million tablets.  

IDC estimates
[Image Source: Statista/IDC]

But it's since slashed its 2016 PC sales prediction to 278.5 million units (a cut of 12.7 percent).  It also slashed its 2015 tablet forecast by a whopping 21 percent.

So we can estimate that between 900 million and 1 billion PCs and between 600 to 750 million ultramobile budget devices (tablets + clamshells, etc.) will be shipped in the next three years.  Most of those devices will be in active use at the end of that period.

Strategy Analytics estimates that in Q1 2015, around 6.6 percent of tablets of the tablets shipped are Windows-based.  IDC in May estimated Windows to hold a 7.5 percent share of tablet unit sales in 2015.  It estimates 9.3, 11.1, and 12.8 percent stakes in the tablet and two-in-one markets in 2016-2018.


That works out to somewhere around 100 million Windows tablet sales over the next three years.  So tablets and two-in-ones will contribute and additional 9-11 percent.

VI. PC Adoption -- Lessons From Windows XP, Vista, 7, 8, and 8.1

That leaves just the traditional PC.  At the close of 2014, Windows market share in the PC space was estimated to be around 91 percent.

Windows market share

A July 2015 estimate by Net Applications pegs the market share at about the same -- 91 percent.  That estimate also indicates that the number of Windows 8.1 and Windows 7 devices has risen at the expense of the now-unsupported Windows XP.
Net Applications
[Image Source: NetApplications]

That same estimate suggested Windows 10 market share prior to launch was around 0.39 percent.  That would indicate an active global population of 1.282 billion PCs, which in turn implies ~1.16 billion internet connected Windows PCs.

The estimate seems fairly reasonable.  That's more or less in line with past estimates, some of which came direct from Microsoft.  In 2011 a Microsoft manager stated that there were 1.25 billion Windows PCs in the wild [source].  Given that PC sales declined in 2012 (3.2%), 2013 (the worst decline by far -- 10 percent), and in 2014 (2.1 percent).  That could cut the total global Windows PC population to somewhere around 1.15-1.20 billion units (vs. 1.25 billion in 2011).

Overall Windows XP -- eight years after its 2001 launch -- reached a peak market share in 2009 reached of around 68 percent of Windows installs, according to Ars Technica.

Windows XP peak market share
Assuming that Windows 10 and Windows 7 split sales 50-50 and cut the roughly 30 percent of devices running Windows XP, Vista, 8, or 8.1 by half, then Windows 7 will likely reach a similar peak market share in the 65-70 percent range.

In terms of new PCs, Microsoft has traditionally has claimed that its popular OS at any given moment occupy around 90-95 percent of new Windows PC sales [source].  But license sales figures suggest this is not quite what it appears.

Roughly 1.01 billion PCs were shipped in 2009-2011.  Thus you could estimate shipments of roughly 855 million Windows 7 PCs. But at the end of 2012 Windows 7 had roughly a 45 percent market share, according to Net Applications (others give similar estimates).

That's roughly 620 million PCs, including upgrades.  The figure agrees fairly well with Microsoft announcement of 630 million Windows 7 license sales in July 2012 (remember, license sales, lead shipments).  Roughly 80 percent of Windows revenue was reportedly coming from OEM licensing at the time and OEMs pay less per license, hence we can estimate that only 5-10 percent of Windows 7 licenses were sold to end users (see all Microsoft's claim that Windows 8 beat Windows 7 in upgrade licensing).  That's about 30-65 million licenses, with an estimated 570-600 million sold to OEMs.

That indicates that the true rate of Windows 7 sales as a percent of Windows PC sales in the first three years was only around 65-70 percent -- far lower than Microsoft's claim of 93 percent.  Microsoft's figure likely only applied to new PC SKUs/models, not older ones -- some semantics magic.  

Ultimately Windows 8/8.1 followed a Windows Vista like pace of sales, if not worse.  But thus far Windows 10 is behaving similarly to Windows XP and 7 before it.  Windows 7 will likely peak sometime this year in market share, six years after launch.  (This places the market share at the halfway mark at approximately 45 percent)

Assuming a return to small growth in the PC space (5-10 percent per year), between 950 million and 1.05 billion PCs will likely ship between mid-2015 and mid-2018.  Recounting the "real" percent of sales of PCs w/ the real sales will probably be around 600 million units.

And by mid-2018 history suggests it should be on at least 45 percent of total PCs (similar to Windows 7 at the three year mark).  Assuming growth to affectively negate the losses of 2012-2014, that leaves about 1.25 billion Windows PCs globally, with about 600 million running Windows 10 (which would in turn give 60 percent of Microsoft's target).  This figure might be assumed in the past to include: minus 60 million licenses on new devices sold in the launch year that have fallen into disuse within three years after launch, plus 60 million licenses for upgrades of new devices.

VII. The Free Factor and Closing Predictions

That all neglects one crucial point.  Where as Windows 7 upgrades were relatively expensive, Windows 10 is free to a large percentage of PC owners, as long as they upgrade within the next year.  Hence upgrade rates should be drastically higher.  

I estimate that at least 1 in 3 Windows 7, Windows 8, and Windows 8.1 users will upgrade given that it's free if you do it within the first year (that's skewed by enterprise users largely not upgrading -- the consumer upgrade rate might be closer to 1 in 2).  Those OS are on roughly 77 percent of PCs -- or around a billion PCs.  Hence in the next year as many as 330 million may upgrade.

Thus when the dust settles I expect the three year Windows 10 market share to be at around 66 percent -- roughly 2 in every 3 PCs.  That's around 850-900 milion units -- about 2/3rds of which are comprised of new devices, and about 1/3rds of which are old ones.

So to conclude, by mid-2018 Microsoft appears on pace to reach:
  • 850-900 million PCs (35% upgrade; 65% new devices)
  • 80-120 million tablets
  • 40-60 million smartphones
  • 45-55 million Xbox Ones
  • 10 million IoT devices
  • 2-5 million HoloLenses
TOTAL: 1.025 billion to 1.150 billion units.

If my prediction is in the ballpark, as many nearly 1 in 4 Windows devices may soon be something other than a desktop or notebook computer (a tablet, a smartphone, an Xbox One, a IoT board, an Xbox One, or a HoloLens).

Source: Twitter [Windows CVP Yusuf Mehdi

"A politician stumbles over himself... Then they pick it out. They edit it. He runs the clip, and then he makes a funny face, and the whole audience has a Pavlovian response." -- Joe Scarborough on John Stewart over Jim Cramer

Latest Blog Posts

Copyright 2017 DailyTech LLC. - RSS Feed | Advertise | About Us | Ethics | FAQ | Terms, Conditions & Privacy Information | Kristopher Kubicki