A closer look at those Windows 8 and Windows RT usage numbers

The latest worldwide usage statistics from NetMarketShare are out. What do they say about Windows 8 and Windows RT? Spoiler alert: Don't believe everything you read.
Written by Ed Bott, Senior Contributing Editor on

It's a new month, which means it's time for our own Steven J. Vaughan-Nichols to launch another attack on the operating system he dislikes so much, Windows 8. This time it's "Microsoft’s New Coke moment."

Unfortunately, his entire argument relies on a handful of significant factual errors.

In the words of Daniel Patrick Moynihan, "You’re entitled to your own opinion. You’re not entitled to your own facts."

So allow me to set the record straight.

First, there’s simple arithmetic.

How bad are Windows 8 sales? In April 2013's Net Applications numbers, Windows 8 barely crept up to 3.82-percent. That still leaves Windows 8 behind Microsoft's last operating system flop, Vista, after seven months in the market.

No, Windows 8 has not been on the market seven months. Windows 8 was released to the public on October 26, 2012. The latest NetMarketShare numbers cover the period ending April 30, 2013. That’s six months and four days.

That kind of detail matters. If someone can’t get a simple fact like that right, should you really trust the rest of their analysis?

Second, my colleague and erstwhile debate opponent continues to compare the launch of Windows 8 with the launch of Windows Vista. He wrote an entire column on the subject in March of this year, creating a graph that purported to compare the two launches using Net Applications data. Unfortunately, the company that actually gathered those statistics has specifically said those comparisons are not valid.

Windows Vista was released to corporate customers in November 2006 and to the public in January 2007. So its first six months on the market would have ended in either May 2007 or July 2007, depending on which starting date you use.

Net Applications changed its data collection methodology in November 2007. The data it collects today is not comparable to the data it collected in 2007.

I asked Vincent Vizzaccaro, Executive Vice President of Marketing for NetApplications, to explain. Here’s what he told me:

Our methodology change occurred starting with data from November 2007. We started doing country-level weighting, which means we compare our traffic to the CIA Internet Traffic by Country table, and weight our data accordingly. For example, if our global data shows that Brazil represents 2% of our traffic, and the CIA table shows Brazil to represent 4% of global Internet traffic, we will count each unique visitor from Brazil twice. This is done to balance out our global data. All regions have differing markets, and if our traffic were concentrated in one or more regions, our global data would be inappropriately affected by those regions. Country level weighting removes any bias by region.

Comparisons of data before and after the change in methodology are invalid because of the massive shift caused by the weighting. It’s apples and oranges now.

We can’t go back before 11/2007 because the weighting required a completely new data collection structure. The two data sets are now incompatible with each other.

For the record, one year after its launch, Windows XP was in use by about 10 percent of the installed base of PCs. I checked the oldest Net Applications figures I have for Windows Vista, which date back to April 2008. Those numbers, which were collected using the same methodology the company uses today, show that Vista was in use by 9.41 percent of the installed base at that time, roughly 15 months after its launch.

And finally, Mr. Vaughan-Nichols takes a shot at Windows RT on his way out:

Windows on tablets fared even worse with touch-screen-based Windows 8 devices and Windows RT devices coming in at 0.02-percent and 0.00-percent each. The last was not a typo. The Surface RT is now in the running for worst Microsoft launch ever.

Oh dear. That does sound awful. In fact, those numbers are so dreadful I asked Net Applications if they’re accurate. The short answer: No. Mr. Vizziccaro explains that Windows RT devices are included with the figures for Windows 8 and are not broken out separately. The entry in the NetMarketShare reports for Windows RT shouldn’t be there.

So, reality check: Windows 8 is following roughly the same adoption pattern as previous Windows versions. The big difference is that its arrival in the marketplace comes as the market for traditional PCs is shrinking. In fact, its most important changes are specifically designed to enable operation on tablets and touchscreen devices, with a special emphasis on mobility.

Six months after the launch of Windows 8, we're just beginning to see the first signs of that shift. According to Strategy Analytics, Windows-based devices accounted for 7.4 percent of the tablet market, or about 3 million devices, in the first quarter of this year. IDC estimates that Microsoft and its partners accounted for 1.8 million devices, or 3.7 percent of the tablet market, in the first quarter. I expect that percentage will increase over the next year. So does IDC, which says if Microsoft can address consumer messaging and cost issues, "we could see Microsoft make even further headway in 2013 and beyond."

When Microsoft launched Windows 8 last year, it also committed to a more rapid update cadence. In the past decade, Windows has been updated every three years or so. The new plan is to release updates on an annual basis. If Windows 8.1 (code-named Blue) arrives this summer as expected, it will be right on schedule.

As for the comparison between Windows 8 and a sugary soft drink, we’re back in opinion territory now. Let’s just say we’ll have to agree to disagree on that one. I do wish that the Coca-Cola Company would release its secret formula so we could see, once and for all, what's really in those classic curved bottles.

Maybe we could even get an open-source alternative to Coke. I'm sure it would take the world by storm.

Editorial standards