Well selling licenses and selling copies of Windows to consumers are 2 completely different things. Selling licenses doesn't tell us anything about adoption and success of Windows. This is why I look at more concrete stats. Any stats you look at, not only the site I linked to, but most of the sites, show Windows 7 absolutely demolishing Windows 8 in consumer usage (aka web and other stuff). And that's month of release not 3 months vs 1 month of Windows 8. There is only one way to count hits from an OS with sniffers really. It was counted the same in 2009 and it's counted the same now. There's is only one Windows 8, there is only one Windows 7.. when you browse you are identified exactly by the OS you use. Even if you are correct and Windows 8 is counted on tablets and on PCs, that would make Windows 8 usage even worse because their stat numbers would entail 2 completely different hardware platforms and they would still be at 0.39%. That's terrible. Only worse is Symbian OS which is pretty much dead these days.
This is a pure spin by Microsoft to try to show Windows 8 as not being a complete failure. And btw, not any kind of evidence, but I can tell you, I know at least 10 people, who bought a new laptop just prior or on Black Friday, it came with Windows 8, and they were so frustrated that a lot of them actually asked or took their new laptop to the computer shops or Geek Squad to have the old WIndows back. Maybe it's just me, but something tells me this is a common occurrence with Windows 8 computers being shipped now.
Let me show you the difference in how stats are counted from 2009 vs 2012.
Here's a graphic. On the left are Q4 2009 results. On the right is the most recent tracking data (updated today) for the same metric. This chart shows all actives operating systems tracked on the web, showing results returned of about 99.75% (I asked it to not include that final 0.25% so that both sides would be equal in length for easier viewing).
Note the huge shift in where web traffic is being handled?
About 97.5% of all measured web traffic in Q4 2009 (around the time that Windows 7 was launched) was desktop/laptop traffic.
The other 2.5% was being handled by mobile devices and Java ME.
Skip ahead to 2012 and now only 90.5% of that traffic is being handled by dedicated desktop/laptop devices.
The other 9.5% is now being handled by mobile devices and the like.
A huge shift in where traffic comes from changes the states on everything else...and makes it redistribute throughout the graph.
This is why comparing Apples & Oranges is never a good idea for proving a point.