Microsoft by the Numbers
techcrunch.com
techcrunch.com
Because no matter how many units Microsoft continues to sell, they no longer set the agenda, and that's the important shift. 10 years ago, the entire industry had to dance around MS on tiptoe. That's over.
There is no longer a single app platform that one company controls that dominates the industry. Microsoft is continuing to milk their golden cow by copying all the best features from everyone else's products and using their sales muscle to push their products in large volumes, good for them.
But they no longer dictate the technical specifications and business models of the platforms I write for, and as long as that's true they can sell a billion smart phones for all I care.
My current company has about 40 linux servers and 3 windows. My previous company had 1 windows server to 5 linux. The one before that was a similar ratio.
I'm looking around online to see how IDC collected/calculated this data but can't see anything that actually says. The only thing I can think of is if they are taking commercial unixes like redhat and suse and comparing them against windows. If that is the case these numbers are beyond worthless due to the fact that most linuxs end up being centos, debian, ubuntu or other free distros.
EDIT funfacts: Here's their source for "linux servers, 21.2%" http://blogs.computerworld.com/15675/idc_windows_dominates_l... which quotes: http://www.zdnet.com/blog/microsoft/behind-the-idc-data-wind... which in turn quotes: http://www.zdnet.com/blog/btl/server-sales-show-signs-of-lif... which quotes this IDC press release: http://www.idc.com/getdoc.jsp?sessionId=&containerId=prU... and here's IDC's description of their information product: http://www.idc.com/getdoc.jsp?containerId=IDC_P348
I couldn't see linux units mentioned there, nor their methodology, or definition of "server".
EDIT2 The bottom of the press release: includes quarterly shipments (both ISS and upgrades) and revenues (both customer and factory), segmented by vendor, family, model, region, operating system, price band, CPU type, and architecture. Doesn't really help, but there's also a phone number and email address.
Because, as others have mentioned, the vast majority of large webhosts or Linux users "roll their own free distro" or purchase an enterprise agreement from RH or Suse, they will never be counted.
Microsoft always plays games with the numbers. They are playing a similar game with Windows 7 numbers - You can buy a Windows 7 upgrade for $99, which only gives you a single license, or you can buy a "family upgrade pack" for $99, which gives you 3 upgrade licenses.
Guess which one allows them to count 3x the number of licenses sold? Right, and they made it so that the family pack can even upgrade pirated versions of XP (I know this because I used this to bring one of my home copies up to legit status). It's a genius marketing move because for basically $33 per computer they get all the pirate computers to become fully legit, counted installs of Windows.
Never mind all the double-counting and triple-counting of licenses - MS has always been king at this. They force Dell/HP to bundle a license of Windows 7 with every computer sold knowing that corporations have Volume License Keys and wipe and reinstall Windows with their corp version. 2 copies of Windows sold for every physical computer.
What I would really like to see are the numbers of Windows computers receiving updates. That would be a legitimate number.
Actually, that number wouldn't be valid either. How many corporations block Windows Update because new versions of IE break some intranet we app?
Not necessarily ... at least not in Europe where I live ... if you're a consultant that gets hired, try explaining to a client that they need to also buy a Windows Server license (or several); or pay extra per instance (in the case of cloud computing like EC2).
Server encompasses a much bigger sphere than 'web-server'. It's in the back room of your local council, the car parts company with a few stores across a couple of states, and all those damn sharepoint installs..
My netbook at home came with Windows XP. It received a full install of Linux the day I laid my hands on it.
They probably count as three Windows users.
A finance firm with over $50bn under management had over 300 Windows servers and maybe a dozen linux machines. Mostly security appliances, firewalls, etc.
A law firm with offices across the globe has over 600 Windows servers and not more then 2 dozen linux machines. Again, mostly security appliances.
Having worked at both these places I can assure you that the licensing fees are a rounding error in the overall expense budget.
That said, I run only linux and mac at home :)
Based on the linked sources from the Microsoft blog post, I'm pretty sure that's what they are referring to.
Also, "server" can mean a lot of things, more than just "box that serves up websites". Fortune 500s often have a giant slew of them just serving email to employees.
From http://www.idc.com/getdoc.jsp?containerId=prUS21703309 I've even more willing to believe that the numbers are BS and are entirely about revenue.
There are a couple of thousand servers, across three DCs, a mix of HPUX, Solaris, and Windows.
Not one Linux server. Maybe there's a few dozen running in VMs on developer's workstations, but even they tend to prefer Solaris, since that's their target, anyway.
At this scale, nobody really cares about the licensing costs of an OS. There's no reason not to buy RedHat, if only to keep the PHBs happy.
OS license costs are a smidgen, a rounding error, damn near nothing compared to the costs associated with standing them up, not to mention development, ongoing operations, etc. I'm pretty sure our storage costs, per month, exceed all the money we've spent on licenses in the last year (if not longer).
I also know of a lot of small businesses which have nothing but Windows. I know of few that have even heard of Linux. I'm talking about every law office, every doctor's office, every real-estate office...well, you get the idea.
I like Linux. I've been running it on servers and desktops of my own for over about fifteen years (since the 1.x kernel). I seldom see it in the field.
Of course they also run Windows, but it's only for Exchange, Windows file servers, and Active Directory for desktops.
Your environment running only HPUX, Solaris, and Windows, is far from the norm. Linux is widely supported, and most importantly, cheap, compared to HPUX or Solaris (running on custom hardware like Itanium or Sparc).
~54% apache + 5% nginx + 1% lighttpd.
Sure, maybe MS won't ever win in servers or smart phones.
Will the PC just die and give way to these?
Not likely, not while most WORK is done with client PCs. How many mid-level bureaucrats are going to be banging away on their IPADS?? five years from now? How much is the client side worth? Lots. WORK? Companies? Sure, big areas that isn't MS has opened up. But the PC is going to be an ocean for a long time and MS, if anything, is cementing it's control of that. I'm no MS booster - I'd like to see a real Linux desktop but I can see MS is better positioned on the desktop than three years ago.
Sony definitely loses money on every console.
Don't know about MS.
If Microsoft had just invested all that money in the bank, they would be a far richer company than they are now.
Of course, the xbox may start returning benefits now, or it may disapper in the next generation.
By late 2006, MS had already lost $4 billion to the XBox project 2 years later this was out to $7 billion. I think recently they've been turning a profit of about $200 million+/quarter
As for XBox Live, it is an impressive service, but the numbers may not tell the whole story. The article cites 23M XBox subscribers, but doesn't say whether these are paying (gold) or free members. I wonder what that mix is.
Even if they're still in the red, the amount they've invested in it is trivial to MSFT. Whatever they spent, it was well worth it to get a web-connected, media-centric device connected to 40 million TVs.
Also there are ovre 40 million consoles sold, so that 23 million must be paying users. I can't imagine anyone having an Xbox without the free service (I'm not even sure you can).
That's not quite true, unless you're talking about games sold per console (which isn't that interesting when you're getting outsold by 30% overall)
Yet another example of ESPN being one of the worst content cartels of them all.
Your ISP has to pay ESPN a monthly fee. I imagine mose cable companies don't, but as noted in another comment AT&T definitely does.
Look at the "smartphone" numbers. The implicit message was that Apple doesn't really have all that much marketshare. Whatever. But, what's Microsoft have? Close to nothing, and there's little evidence that it's going to catch up. No one in his right mind can claim that the "phone" won't replace many (more) current desktop/laptop use cases in the coming years. What slice of that goes to Microsoft?
I don't know much about the CRM space, so I can't comment on the Salesforce.com vs. MSFT thing. But, what's notable is that Microsoft doesn't really enjoy network effects in that space. It must compete like other enterprise software companies -- again, not a model to justify a high P/E. The netbook numbers can be counted as a "save" against consumer-facing Linux distros, but I understand Microsoft to be practically giving away Windows 7 "starter" edition to keep share. Hardly a great model. Furthermore, the OS is little more than a platform for a web browser. Microsoft is doing all the dirty work and making little in return.
What if IBM blogged about its huge mainframe market share in the Fortune 500? Should we be excited about the future prospects of z/OS? How many mainframe customers would gladly get rid of their mainframes if switching costs were lower? 80+%? Look at the number of skilled COBOL programmers out there. Surely such a large development community will continue to propel the platform forward?
What's causing Microsoft's stock to stagnate is that the company has failed to maintain the network effects that fueled its growth and pricing power. They're becoming yet another company competing for consumer nickels and corporate dollars. While that's well and good, it's not the stuff of hocky stick growth curves.
And, Bing? I started using it because of 15% Bing Cash Back deals. Forget doing an actual search or clicking an ad.
Absolutely.
Office isn't going anywhere in most business settings. If anything, the last two Office and SharePoint releases have only continued strengthening its position.
I definitely wouldn't bet that the desktop/Windows/Office thick client paradigm will last forever, but I also wouldn't dare bet against it in the near-term.
Off-topic, anyone know of a good open source web-based office suite project? It would be nice to have the collaboration benefits of web-based productivity software run from a centralized location, combined with the data security of being hosted on private equipment on the LAN.
I feel sorry for the companies who use Sharepoint for its document sharing/management... In time, it will prove a huge competitive advantage for their competition.
My point about extending SharePoint was that you said "I feel sorry for the companies who use Sharepoint for its document sharing/management" and I was addressing that by mentioning a product we use to fill that gap in SharePoint.
There are many reasons for using SharePoint, there's no need for me to get into that here. If you need to add document management to SharePoint, Laserfiche is a nice easy way to do that.
SharePoint integrates impressively well with Word and Excel, for version control and as a data store. Using Office with SharePoint definitely adds value in ways that Office replacements don't currently.
In that way, SharePoint strengthens Office's hold on the desktop, which strengthens the resulting network effect.
To refresh your memory, the article point is that MS is going to be making bank with PCs for a while. End of story (literally - used correctly for once). MS will be selling a gazillion copies of office five years from now.
IBM is still cleaning up with mainframes BTW.
As to whether you should get excited, well I'm not your emotion manager.
Not claiming they are the best tools out, just that they get the job done for me.
You haven't met people using windows 7? Moving away from the tech crowd pretty much everyone I know is using one of xp/vista or 7.
I know Microsoft goes around university campuses trying to get students to use Azure for projects, and hand out a lot of test accounts in the process, inflating their numbers.
Guilty one here. While I use Windows 7, Windows server (multiple licenses) etc at home, I guess just Windows 7 is the norm for most people I know.
At work I can attest to having some hundred Windows Servers and not a single Linux box. As we were a consulting firm, we had lots of different customers but our focus-area was MS-tech. Ofcourse my numbers will be biased, but let me tell you: there are lots of Windows Servers out there. Lots. Sure some Linux-servers here and there as well, but Windows Servers are having pretty good days all in all.
As for Windows Live Mail... I dunno, I like gmail better.
As for Windows Live Messenger though... At first that seems like a really odd question to me, but really, it's not that special.
In my experience IM software is one of the categories of software with the highest geographically variation of any I've come across. In the US I hear people actually use AIM, which is entirely unheard of in the rest of the world. In Europe, you will find MSN is the status quo. In Australia Yahoo Messenger seems to be pretty big (or so it seems), but I've never seen anyone use it elsewhere.
The interesting thing about this kind of software is how it's not the software itself which is drawing users. It's who you can reach (ie, userbase). So if everyone in your area is using product X, it doesn't matter if product Y does everything ten times better, it will still be useless to you unless you can reach the userbase of product X.
Whoever gets users first wins (because who seriously bothers to run several IM applications?). And then it just seems to stay that way.
If I were to take my (severely) biased impression and generalize, I would say everyone is using every product from Microsoft, with the noteworthy exception of BizTalk. Linux can only be seen on netbooks or as a means of bypassing the corporate web-filter. Oracle does exist, but only in very few companies which rakes in billions. MySQL doesn't exist at all, except for in the bottom-tier web-hosting basket.
But I'm not going to do that, and I suggest you don't take (what I assume is) your Linux-hacker biased experiences and overgeneralise them either. The truth is probably somewhere in the middle.
Not everywhere. Mostly ICQ here in Germany.
(links to http://technologizer.com/2010/06/27/seven-copies-of-windows-...)
When they came out with any software, they chose the most confusing name possible. "Windows" was a generic term for any the panes in windowing GUI, until MS appropriated the word. Then came "Word", which sounded like about all the competition out there. Once they had the OS market sewn up, they used that to make their other software work better than anyone elses, with undocumented APIs. Then they offered "bundles" which meant, if you used their OS, which of course every business had to, you couldn't afford to turn down their other offerings. This might sound like fair, tough competition, but the reality was that they used their money and clout to run the competition out of business, then their de facto monopoly to make a fortune.
And tough but fair, often gave way to threats. If a retailer didn't play ball, getting rid of competition, they would pull their joint marketing/advertising deals. If a small business came up with anything interesting, they'd swallow it, or run it out of business by any means necessary. This did more to stifle innovation for about 15 years than anything else.
This is only the beginning of why people don't like MS. If we were to all post here, it would become an encyclopedia.
You can inform yourself.Search for DRDOS.Search for windows trademark issues,lindows, you know what? they were an entire line of windows branded products before MS, MS destroyed them all, you can search what happened to them.
Windows is a generic word, it can't never be trademarked(unless you are rich to get over the law), that applies to "word", "powerpoint", "project", "excel", "exchange". A word in the English dictionary just can't be trademarked, by law.
Maybe you were a kid then, but there was a time when "word" was not the most used word processor, it was "WordPerfect", and people used Lotus123 instead of "excel". What did MS did? They made windows but didn't let WordPerfect and Lotus123 people(and everybody else, like compiler builders) use the windows API, so MS had a 4 year period of advantage. Once they did, the high level exposed API was slower than what MS used.
I'm tired too. When people don't know they don't know what they don't know.
Windows is an English word, however that doesn't mean it doesn't enjoy some trademark protection in an arrangement like "Microsoft Windows". The protection is not as significant as it would be on an invented word (like, say, "Microsoft") but it's still there if you come up with something that's judged to be similar enough to cause confusion.
Similarly, try founding an IT company called Apple and see how far you get - that's an English word too.
IIRC Gary Kildall special on The Computer Chronicles show mentioned the case when DR agreed with IBM to bundle CP/M-86 on IBM PCs only to discover IBM set the price of MS-DOS at $40 and CP/M-86's at $240...
If nerds ever tear down MS's walls, it will be well deserved.
http://en.wikipedia.org/wiki/United_States_v._Microsoft
http://grokdoc.net/index.php/Dirty_Tricks_history
http://www.technologyevangelist.com/2007/02/microsoft_dirty_...
Another example, they built the browser into the operating system, thus killing of Netscape. Once they did that they let IE stagnate for several years, providing no innovation (IE6 is 2001). We are finally crawling out of that hole. IE9 is going to be good because it has competition and Microsoft now must compete or see further erosion in their browser market share.
In short, because Microsoft has a desktop monopoly (90% for 15 years), and they've used it aggressively, they make people a bit unhappy. Personally, I don't care if people choose to use Windows, but many people don't want to, yet they must because of "corporate standards."
The only other thing that bugs me is people always want the govt to look into Apple and Google. Everyone seems to have forgotten that Microsoft still has this huge 90% market share. If it could be reduced to 70%, for example, everyone would benefit, including Microsoft users. What's the reason to go after a couple of "small" fish nipping at the heals of Microsoft?
"The only problem with Microsoft is they just have no taste. They have absolutely no taste. And I don't mean that in a small way, I mean that in a big way, in the sense that they don't think of original ideas, and they don't bring much culture into their products.... I am saddened, not by Microsoft's success — I have no problem with their success. They've earned their success, for the most part. I have a problem with the fact that they just make really third-rate products."
Their hardware is not shabby either...zune, xbox 360 (well hardware quality aside), mice, keyboards.
IE6 is not a mistake. It's old. It first come with Windows XP and Windows XP is 8 years old. At the time, IE6 was fine.
The problem arose from Microsoft disbanding the IE team and not releasing a new version for so many years. It fell behind from age.
Do Microsoft have an Intranet focused IM solution ala Lotus Sametime? Notes shops all seem to be running sametime, Exchange shops seem to be not running IM or running public Messenger.
It's possible that the reason MSN is so popular with my age group is that we grew up with it. MSN was completely tied in with hotmail which was the email service of choice for most of us when we were getting our first email account in the early 2000s (age 9+). Then again my 8 year old cousin also uses MSN.
That said, the only reason I'm trotting out that installation is to test my sites on IE. Otherwise, It's Ubuntu all the way. Probably Android when Intel releases the Froyo build. Also maybe Chrome, just for fun.
In other words, the only reason I'm keeping Windows is I like playing with barely functional software. (I kid, I kid. Though it did hang for 4 hours on the logout screen.)