“We all know what its like to imerse one’s self in the age old computer market-share debate. The problem is that when most people use the word “market-share”, what they really mean is “installed-base.” Because consumers choose a platform based on its popularity, misconceptions about market-share only confuse the issue even further.” Read the editorial at OSCast.
OS Market-share vs. Installed-Base
About The Author
Eugenia Loli
Ex-programmer, ex-editor in chief at OSNews.com, now a visual artist/filmmaker.
Follow me on Twitter @EugeniaLoli
It makes sense, but I thought everyone already knew to take statistics with a grain of salt. Which, of course, explains the endless debate in these forums about the subject. ; )
Well at least 4 out of 5 people see it that way.
no matter how you cut it, the majority is still the same
today, from the top end to the low down, windows pretty much covers the whole personal desktop spectrum
this weekend, I dig out a Pentium 133 notebook with win98se. To see how it will do with the latest office suite, I loaded IE6 and Office XP on to it. With 48 MB of EDO RAM and a 2 GB hdd, it works like a charm – apps will open in 3 to 8 seconds – definetely put latest linux desktop to shame with bloated apps like OOo, mozilla, etc.
The author has interesting points buts he clouds them with his lack of organizationation or a bottom line (i.e. a thesis). It is really unclear who he is criticizing here. His article needs a follow-up that explains why he needed to write the article.
The problem is not with the statistics themselves but with their interpretation through the media. Thus, the problem lies squarely at the feet of this editorialist. Personally, I thought it was a great article and I learned a lot from it.
So where does his date come from? The 3-5% market share figure is easily obtainable. But how do you determine the installed base figure? Is there somewhere a report online?
What it doesn’t take into account is the different uses of old hardware. People point to things like google’s zeitgeist, which shows an overwhelming use of Windows and IE. Those people who are “squeezing the life out of” hardware with Linux/*BSD aren’t using those as desktop systems, for the most part. They sit in closets as servers, etc. etc. While there might be something that’ll run on them and be usable as a desktop, there’s little reason to do so. I *could* run something like a minimal WM on my old SS10, along with Netscape4, etc, etc. It does much better as a generic X terminal running other things, or as a dedicated server.
On the Mac side, I don’t see a whole lot of System7 machines around in use anymore. The few that I have seen do things like MIDI sequencing, using old programs, and are pretty much dedicated to those tasks. But machines running OS8 or better I see all over the place, just as I see Win95 machines.
And then there’s the Amigas running Video Toaster. 😀
“So where does his date come from? The 3-5% market share figure is easily obtainable. But how do you determine the installed base figure? Is there somewhere a report online?”
“Naysayers have been calling for Apple’s demise for years. But Apple not only has survived but thrived, it seems, at least partially by the sheer force of Jobs’ will and his ability to maintain the ferocious loyalty of Apple’s users, who still account for 10% of the world’s computer users, while its sales usually account for about 3% to 5% of the world global PC market.”
http://www.forbes.com/2003/06/02/cx_ah_0602jobs.html
“The author has interesting points buts he clouds them with his lack of organizationation or a bottom line (i.e. a thesis). It is really unclear who he is criticizing here. “
Sounds to me like he just has a beef with people saying market share when when what they really mean is installed base… and equate percentages that are misleading.
As the article says, “Becausee consumers choose a platform based on its popularity, misconceptions about market-share only confuse the issue even further.”
His bottom line or theses appears to be that he wants people to not feel intimidated to buy into a computing platform based on ilconceived notions that its installed base (mistaken as market share) is inadequate to make the investment worth while.
he makes a valid point.
If Apple and Linux’s market share are in the 12% range, they are both extremely viable.
“What it doesn’t take into account is the different uses of old hardware. People point to things like google’s zeitgeist, which shows an overwhelming use of Windows and IE. Those people who are “squeezing the life out of” hardware with Linux/*BSD aren’t using those as desktop systems, for the most part. They sit in closets as servers, etc. etc. While there might be something that’ll run on them and be usable as a desktop, there’s little reason to do so. I *could* run something like a minimal WM on my old SS10, along with Netscape4, etc, etc. It does much better as a generic X terminal running other things, or as a dedicated server.
On the Mac side, I don’t see a whole lot of System7 machines around in use anymore. The few that I have seen do things like MIDI sequencing, using old programs, and are pretty much dedicated to those tasks. But machines running OS8 or better I see all over the place, just as I see Win95 machines.”
Its interesting that you brought this up because every PC sold, whether its being used as a server or a desktop is factored into the “Windows PC” market share statistic. First of all, It doesn’t matter if that computer was later wiped and had Linux installed on it, its still a “market share” tic for Windows because it was sold with Windows.
Additionally, because market share doesn’t calculate differences between server, office desktop, or consumer desktop (instead just simply being a Windows market share point) the figures get distorted even further.
This guy is really on to something. Lets not be influenced by the incorect market share statictics anymore, as they inacurately favor Windows.
The author states that the Mac installed base (for desktop systems) is around 10% as is that of Linux, leaving Windows with an 80% share of the installed base.
Baloney! Actually the figure for Mac is just a few percent at most and for linux and unix less than one percent. This leaves Windows with a 95% share of the installed base as well as market share, at least. (The real market share for MS could be even higher in the near future because of expensive subscription licensing).
While the author has a good point about the difference between installed base and market share, he was quite deceptive in sneaking in the assumption that linux has 10% of the installed base. Figures from Google showing the operating systems of origin in making queries there and from other independent sources whittle down the installed base for linux considerably from what linux boosters would have us believe. There are only so many geeks in the world, and even if every geek in the world were to use linux as a dekstop system that wouldn’t put linux above 1% of the installed base. And I seriously doubt that Apple’s OSX has a full 1% either. Probably another 1% are using older versions of Mac, Amiga, etc., still leaving Micrsoft with 97% of the installed base of desktop systems, easily.
I don’t know about the US but in Australia we have vast numbers of old Volvos on the road. The installed base of (10-40 yo)Volvos in Australia is huge. Is that great for Volvo – not really – they aren’t serviced by Volvo factory dealers or using genuine Volvo spare parts in many cases. People aren’t trading their old Volvos for new Volvos very often. In addition people think crappy old Volvo station wagon when the name Volvo is mentioned not new turbocharged all-wheel drive models.
Substitute Volvo for Apple and installed base isn’t that great a deal.
Aw man. That was a hilarious comment.
“This book can’t be wrong–look, it says on the cover that it’s right!”
I’m sure you have conducted a huge amount of surveys to yield your information for “Hmm, I think maybe 2% of people use Apple. And no more than 1% use Linux. Therefore MS has 97%.” Great reasoning bucko.
“While the author has a good point about the difference between installed base and market share, he was quite deceptive in sneaking in the assumption that linux has 10% of the installed base. Figures from Google showing the operating systems of origin in making queries there and from other independent sources whittle down the installed base for linux considerably from what linux boosters would have us believe.”
Your theory assums that every one of these computers are being used to browse the internet. Of course, many Linux and Mac Personal Computers are simply used as work horse machines and not used to browse the web
“Substitute Volvo for Apple and installed base isn’t that great a deal.”
Except for the fact that these Mac users are buying software… a factor which these arguments typically stem from.. (that developing for the platform wont result in enough ROI because of Apple’s “3%” market demographic.
If Apple’s install base is 10-12 percent…. there goes that theory.
I don’t think you’re replying to the right person.
Also, I do not understand what you’re talking about.
“The author states that the Mac installed base (for desktop systems) is around 10% as is that of Linux, leaving Windows with an 80% share of the installed base. Baloney! Actually the figure for Mac is just a few percent at most and for linux and unix less than one percent.”
Sounds like somebody didn’t fully grasp the article. You’re touting market share rather than install base.
“This leaves Windows with a 95% share of the installed base”
You mean market share.
“The real market share for MS could be even higher in the near future because of expensive subscription licensing).”
Now you’re mising profit margin with install base AND market share. (You’re forgetting, these people already bought the computer… they’re not increaseing the install base.
“While the author has a good point about the difference between installed base and market share, he was quite deceptive in sneaking in the assumption that linux has 10% of the installed base.”
Actually, those figures are probably pretty acurate considering the fact that Linux’s 3% market share is determined only by Linux Sales which ofcourse is a misnomer because most people dont buy Linux.
“Figures from Google showing the operating systems of origin in making queries there and from other independent sources whittle down the installed base for linux considerably from what linux boosters would have us believe.”
This assumes that each of these computers is browsing the web. This isn;t the case, as many of these machines are likely to be dedicated workhourse machines that have little to no need to browse the web.
“There are only so many geeks in the world, and even if every geek in the world were to use linux as a dekstop system that wouldn’t put linux above 1% of the installed base.”
Perahps, but then again, you totally lost any grasp of the issue at hand with that statement.
“And I seriously doubt that Apple’s OSX has a full 1% either.”
You’re probably right, and yet there is a total 12% mac user base.
“Probably another 1% are using older versions of Mac, Amiga, etc., still leaving Micrsoft with 97% of the installed base of desktop systems, easily.
Again, you’ve totally mi-understood the issue at hand.
Sorry Greg, that comment was meant for annonymous.
“Except for the fact that these Mac users are buying software… a factor which these arguments typically stem from.. (that developing for the platform wont result in enough ROI because of Apple’s “3%” market demographic.
If Apple’s install base is 10-12 percent…. there goes that theory.”
I doubt that users of Macs older than about 3 years are buying much software at all. If you can’t run OSX you won’t buy it or upgrade your apps to use it.
Well at least 4 out of 5 people see it that way.
Gah! More statistics!
Where did Forbes get the data? And where did the article get the data for Linux’s install base?
Thin air?
40% of all statistics are made up, 60% of the population knows that 🙂
People point to things like google’s zeitgeist, which shows an overwhelming use of Windows and IE.
People need to take google’s zeitgeist with a grain of salt as well. Referring to google’s zeitgeist as “correct” statistic is like saying of course all desktop computers have internet access, of course all desktop computers access google of course equally often and the information about OS and browser is of course always true. Neither of them is actually sufficiently the case in reality, so we can restart from the very beginning regarding statistics.
“The author states that the Mac installed base (for desktop systems) is around 10% as is that of Linux, leaving Windows with an 80% share of the installed base. Baloney! Actually the figure for Mac is just a few percent at most and for linux and unix less than one percent.”
Sounds like somebody didn’t fully grasp the article. You’re touting market share rather than install base.
———-
I’m not going to respon point by point because your whole argument is based on the same assumption the author made, that the installed base of linux is somewhere near 10%. Wrong!
The 10 to 12% of installed base for linux and the 10 to 12% for mac alleged by the author is so far off that it can only be interpreted as deception, and/or humoring linux and mac boosters who want to believe that acceptance of their systems is more widespread.
The only thing presented to counter my point that installed base of all non-MS desktop systems, total, is less than 5% is that many linux boxes are used as firewalls, servers, etc., sitting in closets, not as desktops. Those units don’t count. We are only considering the installed base of desktop systems (including notebooks) for end users – machines on which linux is the primary desktop in use. This figure is less than 1% for linux and unix, and not much greater for Mac. If either Linux or Mac had anywhere near 10% MS would not have the monopoly it now enjoys.
It just so happens that Microsoft’s market share and installed base are about the same – over 95%. That wouldn’t be the case if Windows were free, like Linux. Even with rampant piracy of MS software, far more Windows licenses or preinstalls are sold than boxed sets of linux are sold, in nations where piracy of MS software is commonplace. People don’t want linux as a desktop system but it seems that they will be forced to use it in nations where use of free software is mandated by law. But will these laws be enforced?
More onerous forms of licensing will increase market share for MS by forced upgrades of both hardware and software, especially with upcoming DRM technology that requires compatible software and hardware to run.
“Probably another 1% are using older versions of Mac, Amiga, etc., still leaving Micrsoft with 97% of the installed base of desktop systems, easily.
———
Again, you’ve totally mi-understood the issue at hand.
Evidently I have. The real issue is how anyone could believe that linux has anywhere near the percentage of installed base claimed in the artice and accepted without question by linux well-wishers.
I can proove that right off my desktop, not even moving my lazy arse..:
Do you people have friends? Yes, many.
Do you people know many other people? Sure, loads.
Do you people know many individuals having computers? Sure enough.
So, as for myself, I know “millions” of people having “millions” of computers, amongst which I can even count quite a few geeks. 10 + 10 % means on an average every 5th person I know having a computer or being involved with a computer would not be on Windows — and claming that is just plain and outright wrong and lunatic. I couldn’t even thinks of anyone I know who would make such a bold statement. But if we all don’t have every 5th friend not running Windows, then somebody has gone astray here to some extent, I am afraid….
Whilst I type this on SuSE, I am not even sure as what my Mozilla will be recognized on the web, many things have been said about these stats, so I don’t trust them. Still 10% is hilarious. If Linux and Mac have 10%, then BeOS has 5% *lol*
“40% of all statistics are made up, 60% of the population knows that :-)”
Yes, but only 58.735466299% ( approximately ) of them will admit that.
So, as for myself, I know “millions” of people having “millions” of computers, amongst which I can even count quite a few geeks. 10 + 10 % means on an average every 5th person I know having a computer or being involved with a computer would not be on Windows — and claming that is just plain and outright wrong and lunatic.
What if you lived in Asia?
“What if you lived in Asia?”
Then you have a 98% chance of being Asian?
This is spin, pure and simple. Mac has 10% share??? And Linux has about 10% installed base? That’s saying that 1 out of every 10 computer users has Linux installed somewhere! Are you kidding me??
Here’s a clue:
Installed base = roughly the total market share in past years.
Care to elaborate what your question is about? So what, if I lived in Asia?? You mean that the percentage of Macs is higher there..? Oh, are you refferring to the very limited number of Japanese? Well, Japan is not “Asia”, China is a lot more like Asia with far beyond 1 billion citizens. It is not relevant because there are so much Chinese – it is relaveant in this context because the very largest part of folks there and in even poorer asian countries “have never touched a computer” before, not to talk about fancy Macs. The Macs in Asia, although relatively high by % maybe in Japan are not relevant overall and _certainly_ not that relevant so that they would help the overall Mac share to be 10% world-wide. So, that is if you lived in Asia.
Be clever people. Just because you don’t know many people who run a Mac or Linux doesn’t mean it has a low install base (firstly in was a non-MS/Mac OS that made 10% – not specifically Linux).
First off, look at schools, the majority of schools run Mac networks not Windows. The company I work for (a newspaper) runs about 9-10 Macs vs Windows (the only people who run Macs are secutaries). Also, half of my department run Windows at home but nearly all of them (minus myself) run Macs, I run Linux. So if I viewed things only in my little world I cannot believe that MS has an install base of 80% because my personal experience.
At my girlfriends University the vast majority of computers are Unix based (mostly Linux), the only Windows desktops are the odd laptop here and there. Same thing at the bluechip company my Dad works for.
So just because you’re tiny little inexperienced worlds only run Windows doesn’t mean that’s all there is.
As they say “Just because you live in France doesn’t mean everyone, everywhere only speaks French”.
rajan r writes:
Gah! More statistics!
There are three sorts of people, those that believe in statistics and thoses that don’t.
🙂
When Apple counts the numbers of x86s vs Apples sold, they do not include business sales, nor do they include online sales. Considering that Dell and Gateway sell almost exclusively online, and they are two of the biggest PC manufacturers, this is a very big deal. HPQ and IBM also sell a significant of PCs online.
What Apple does is: get statistics from stores like compusa, which sells both apples and x86s. This is grossly missleading.
Home Desktop Systems
Mother: Windows 2K
Sister: Mac OS 10
Brother in Law: Mac OS 10 (2x)
Me: Mac OS 10 (2x), Linux
Supervisor: Windows 98
Friend 1 at work: Windows ???
Friend 2 at work: Windows ???
Friend 3 of family: Mac OS 10
Count: Windows 4(33.33%),Mac OS 10 7(58.33%), 1 Linux (8.33%)
With out leaving my desk; I did a quick survey of friends and family and wouldn’t you know… Mac OS has a better market share then Windows.
With out leaving my desk; I did a quick survey of friends and family and wouldn’t you know… Mac OS has a better market share then Windows.
You mean install base right?
Care to elaborate what your question is about? So what, if I lived in Asia?? You mean that the percentage of Macs is higher there..? Oh, are you refferring to the very limited number of Japanese? Well, Japan is not “Asia”, China is a lot more like Asia with far beyond 1 billion citizens. It is not relevant because there are so much Chinese – it is relaveant in this context because the very largest part of folks there and in even poorer asian countries “have never touched a computer” before, not to talk about fancy Macs. The Macs in Asia, although relatively high by % maybe in Japan are not relevant overall and _certainly_ not that relevant so that they would help the overall Mac share to be 10% world-wide. So, that is if you lived in Asia.
Sure.. First you are bitching about Linux, then suddenly you jump to Macs because it suits you better.. If I said “what if you were a graphic designer”, you would probably be bitching about Linux..
What about browsers that are set to identify themselves as IE 6 when they are actually Opera or some other browser? Was this taken into consideration? Some web sites only display properly or won’t display at all unless the browser idenfies itself as IE 6 or Netscape x.x…
Why? Why is market share so important. I don’t check for market share for any other product. I buy it because I like it. The effects of marketing are wasting your brain cells!
“Why? Why is market share so important. I don’t check for market share for any other product. I buy it because I like it. The effects of marketing are wasting your brain cells!”
It never made sense to me either, but several people feel this overwhelming urge to follow the crowd. Even if a product is significantly better, many fear going outside the norm because they fear that they might be wrong.
If people understand that more individuals bought a computer than what was previously esimated, then that platform gets its due when an individual feels more secure about moving to that platform.
I thought it was a great article.
Let’s assume you and the many Mac shills are right – and Macs have longer useful lives than PC’s (I don’t hate Macs I use them BUT I am tired of this emoting from from otherwise perfectly reasonable techies over a good but ‘also ran’ platform). Let’s also assume that the ‘installed base’ is different from the marketshare (although I don’t see what this assumption is taken at face value by so many – afterall the industry estimates are derived from the application of statistical methods that would have taken the installed base into account as well).
Based on both presumptions above the following MUST be true.
Since Mac’s have a longer ‘useful life’ than PC’s the purchase cycle for PC’s is more rapid (some say three years PC five years Macs). This means that within ten years there would be three PC purchases made as opposed to two Macs – since even the most rabid of Mac fans cannot reasonably assume that all PC’s are rendered useless within three years it is obvious that a large amount of used PC’s survive through the three year re-purchase cycle. WHICH MEANS THAT THE INSTALLED BASE OF PC’s WILL BE ALWAYS INCREASING RELATIVE TO Macs. In fact the only real conclusion that could be drawn from this competely specious and artificial distinction between ‘installed base’ (meaning used computers) and marketshare is that the marketshare picture would be better for Macs than the ‘installed nase’ picture. In short there would be even less Macs to PC’s than the industry marketshare shows.
Let’s be real people – the phenomenon of Apple selling product at record rates while losing marketshare is set to continue – the bad news is that it will only accelerate as the stats are against Apple – since the sheer number of PC’s sold per unit time means that Apple cannot realistically produce and sell machines at a rate to even hold onto their current marketshare – its a losing battle folks and not so hard to understand. So stop making up spurious arguments – Sheesh!
>> This means that within ten years there would be three PC purchases made as opposed to two Macs – since even the most rabid of Mac fans cannot reasonably assume that all PC’s are rendered useless within three years it is obvious that a large amount of used PC’s survive through the three year re-purchase cycle.
Completely agree, so that expains the alternate OS (eg Linux/BSD etc.) 10% as those old machines are recycled into Linux desktops etc.
The argument is between Macs and PC’s not Macs and Windows – mentioning Linux in this case is a red-herring. Sorry.
What about browsers that are set to identify themselves as IE 6 when they are actually Opera or some other browser? Was this taken into consideration? Some web sites only display properly or won’t display at all unless the browser idenfies itself as IE 6 or Netscape x.x…
That doesn’t change the way your OS is reported to the site does it? If I run Linux and set Firebird to identify itself as Internet Explorer does that show in the website log as Internet Explorer on Windows? I didn’t think so, but I may be wrong. So changing your User Agent to show Internet Explorer shouldn’t scew the numbers for install base of an OS.
This article does not seem credible. The author is coming to some very interesting conclusions. If we had some sources I might buy the Mac statistic, although from what everybody says about Linux these days it seems like you need a fast computer for the newer linux distros just like you need a fast computer for Windows 2000/XP. But otherwise it just sounds like the author pulled numbers out of his . . . . .
Actually, I beleive that the web server depends on the browser to identify the OS, so if your alternative browser identifies itself as IE/Win, then google Zeitgeist counts another one for IE/Win.
I think you lack real news, but thats the world somethimes, borring….just joking
“Figures from Google showing the operating systems”
….tell me i use IE even though i can hoestly say that i have never in my life used it even once. not on any of linux machines or on my mac. but they all think i am, otherwise i can’t use online banking. go figure.
also, i really like the idea that i can figure out the worlds user base from my friends, so i did a little thinking on this. turns out that 11% use mac OS X, 5% use 9.2 or earlier, 34% use linux and everyone else hates their computer. wow!
also, 68% of the population is female, 14% gay, and only 3% black. oddly 2.5% of the population is named “jenn.” everyone on earth lives in the USA or japan. oh, and there are more computers than people.
I will defer to Eugenia on this although I seldom defer to her instinct to “just know what is right”, as she says.
What is your best estimate of the actual installed base of linux desktop systems and macs? In articles months ago you quoted figures for Linux at less than one percent. That figure is confirmed by statistics on query origins at Google and by other surveys made by industry analysts.
On the other hand a few people posting here swear that among those they know, all of whom can’t be geeks, use of Mac and Linux exceeds use of Windows at home, and that similar informal surveys of friends, relatives and associates would show much higher usage of Linux and Mac than is acknowledged by the “mainstraim” industry.
I still feel that the 10% figure for linux is way too high. However, I’m certainly hoping that linux can break the Windows monopoly, gaining 10, 25 and even 50% of market share but not 95% like Windows. Why not total domination? To make room for other open source operating systems more oriented towards home users and small business, which are not knockoffs of unix (yuck!) like linux is. For example, OpenBeos and other spinoffs.
What is your best guess of installed base for desktop linux as a percentage of total intalled base and have you seen much growth in the percentage (not absolute numbers) in the past 2 or 3 years? Of course there has been growth in absolute numbers which is meaningful because even if the percentage is unchanged (due to more people also getting Windows boxes) that gives linux a bigger user base and is good for continued development. However, that’s a different issue.
John
The argument is between Macs and PC’s not Macs and Windows – mentioning Linux in this case is a red-herring. Sorry.
Yeah Lee, my ‘friend’, get the argument right. It’s between OS’s did you read the article or are you as stupid and niave as your comments make out?
Here’s a tip, read the article, in fact just try reading the title of the article “OS Market-share vs. Installed-Base”, nowhere does it say ‘Mac vs PCs’ especially seeing Macs can run Linux and other machines such as Sparc stations are run as desktops.
Just a question, what sort of moron joins a discussion board when he hasn’t even read the title of the discussion?
you must consider that most mac and linux users also have a windows box at work or for some other reason, but they still consider themselves mac/linux people. Myself included, and most of my friends that use an alternate OS
I dont think apple has 12% user base. I think it is around 5-8%, but that 5-% is not shrinking, and is a very active user base.
when the marketshare number is shrinking, that doesnt mean people are switching to windows either. It is just part of the fact that the computer market is still growing, and some user bases arent growing as fast as others.
Only 5% of the people on the planet are cool. God made me cool.