Windows 8's sales are worse than Windows 7's were at the same time in its lifecycle—unless they're higher. They're also apparently worse than internal Microsoft estimates. But what those estimates are is anyone's guess. Truth is, we don't really know and we won't for some time. But combine the new operating system with the continuing conjecture over former Windows Division president Steven Sinofsky's departure, and you've got some great headlines.
The first mutterings about poor sales came in the aftermath of Sinofsky's demise, with incredulity that he'd be let go so soon after Windows 8's launch unless sales were catastrophic. Bad sales are plausible, too, after reviews of Windows 8 were decidedly mixed. The operating system, at best, seemed deeply polarizing.
Surprisingly, the first talk of sales numbers was positive: the Register reported the results in Europe were, according to channel analyst Context, actually quite good. Sales in the two weeks up to and including the October 26 launch of Windows 8 were up 7.8 percent year-on-year, and about a quarter of these machines shipped with Windows 8. Windows 7 was only installed on 17.1 percent of machines sold in the equivalent time period around its launch. Context also noted that in spite of fears to the contrary, there wasn't a glut of unsold Windows 7 inventory clogging the channel. In a sluggish European computer market still suffering the effects of the global downturn, that's a pretty healthy performance.
A few days later, the news wasn't so good. Microsoft-watcher Paul Thurrott wrote that a single source inside the company said Windows 8 PC sales were "well below" internal projections. The software maker apparently blamed the slow start on lackluster offerings from PC OEMs.
Today, AllThingsD offered confirmation of sorts to Thurrott's source. According to market research firm NPD, the US market does have a problem with stale Windows 7 inventory. Windows 8 is growing each week as these old machines are flushed from the channel. But Windows 7, in contrast, had no comparable problem with Windows Vista inventory. The result? Windows 7 had a "stronger launch," but Windows 8 has a "faster ramp."
Microsoft has announced no sales figures. But Redmond did say that over the first three days of retail availability the company sold 4 million upgrade licenses and the operating system was selling faster than Windows 7. While the time scales may be different—a few days versus a few weeks—this would appear to undermine NPD's claim that Windows 8's launch was weaker than Windows 7's.
Even Thurrott's source gives no real insight into sales figures. Windows 8 could have matched or beaten Windows 7, but still fallen short of Microsoft's own projections if those projections were high enough. At its earnings call in October, the company pointedly refused to give investors any guidance on its expectations of the Windows division's revenue even when directly asked.
If the company does have projections, and it surely must, it doesn't yet have enough confidence in them to issue guidance based on them. If we're cheeky, we might, however, be able to speculate a little from its other predictions—in particular the relationship between the Microsoft Business Division's Office revenue, the PC market as a whole, and Windows revenue.
Microsoft splits Office revenue into two parts. There's multiyear revenue, from volume licensing subscriptions, and there's transactional revenue, from one-off perpetual sales such as the various editions bought at retail or pre-installed on PCs. In general, Microsoft expects transactional Office revenue to reflect the performance of the overall PC market, albeit at a slightly lower level. We can see how this would make sense: a growth in sales of consumer and small business PCs will cause a corresponding growth in sales of consumer and small business versions of Office (the ones that tend to be transactional rather than subscription).
Windows revenue, particularly transactional Windows revenue, is similarly driven by the overall health and behavior of the PC market. This tends to follow in lockstep for reasons that should be obvious.
For the three months of October, November, and December 2012, Microsoft expects transactional Office revenue to grow by the low single digits. From this we could infer PC market growth of perhaps mid-to-high single digits, and hence Windows revenue growth of the same amount.
Although only spanning two weeks, the European numbers from Context are in line with this inference. The numbers from the US may be worse—given the contradiction between Steve Ballmer's claims and NPD's, it's difficult to draw any coherent picture. But this inferiority may be due not to poor sales per se, but delays in shifting Windows 7 merchandise. With NPD noting the old inventory is being cleared and Windows 8 is ramping up, the US could soon match Europe's results.
Absent an explicit statement on sales figures from Microsoft, the true impact of Windows 8 on the PC market and Microsoft's own finances won't become apparent until January. Then, we'll get real numbers for this quarter's performance.
Nonetheless, it's far too early to write off the operating system yet. The evidence, such as it exists right now, is nowhere near strong enough to dismiss Windows 8 as a catastrophe. Windows 8 may still bomb on the market and its success certainly won't be straightforward or obvious given the nature of the changes that it makes. But its failure is by no means a foregone conclusion either—and it might just surprise the critics in the end.
reader comments
196