Is Linux Commoditizing the Desktop OS Market?
Glyn Moody wrote an interesting article for the Guardian earlier this month titled “Why falling Flash prices threaten Microsoft.” It got me thinking about commoditization; specifically pondering the question: Is Linux driving the O/S towards commoditization as many would have us believe?
The key point of Mr. Moody’s article is that several converging market forces are poised to impact Microsoft’s dominance of the operating system and productivity software markets. Those forces are as follows:
- low hardware cost of the new ultra-portable computers such as the Asus Eee
- low memory and storage requirements of many Linux distributions
- increasing memory & storage requirements of Windows distributions
- emergence of free web-based productivity applications
While the author briefly mentions the usability of the Asus Eee as being responsible for its success, his primary argument revolves around the cost differential between the Linux and Windows versions. He rightly points out that while consumers are being drawn to web-based applications such as Gmail and Google Docs, and are spending an inordinate percentage of their computing time using a web browser (all of which require very few client-side physical computing resources as opposed to traditional desktop apps); each new release from Microsoft requires more and more physical resources from its host system. These resources are needed to support a glut of features that the “everyday user” finds unnecessary, and for which they are unwilling to pay a significant premium. As the premium for Windows systems increases as a percentage of the price of alternative Linux systems, Windows will lose market share.
I hope that I’m not putting words in his mouth, but it seems to me that Mr. Moody is saying that the operating system is on the road to commoditization and that, as a commercial product, Microsoft Windows will not be able to compete with its free open source alternatives. While I agree with all of the noted observations, I don’t quite buy the demise of Microsoft as a conclusion. Now, as representative of an open source advocate, I know that I’m treading on thin-ice if I defend Microsoft, but bear with me for a moment….
The simple fact of the matter is that when it comes to desktop computing, the operating system is absolutely not a commodity. Here are some numbers to back-up that statement:
- Microsoft reported a 15% growth in revenue from 2006-2007. Much of this growth is attributed to sales of Office and Windows Vista (over 100 million copies of Vista sold in ‘07).
- Microsoft reported $14.8 billion in revenue from Windows licensing in 2007, 80% of which came from their OEM channel.
- Apple reported $10.3 billion in revenue from sales of Macs in 2007; another $1.5 billion in software sales, including their new UNIX-based Leopard operating system.
If the operating system market was commoditized, then by definition cost would be the primary point of competition. As we can see, more than a few consumers and businesses chose costly, proprietary operating systems from Microsoft and Apple over free Linux alternatives. I don’t believe that a stronger argument need be made against the thought that the operating system is commoditized.
Instead of commoditization, what we’re seeing in the ultra-portable market is the realization of a viable niche for Linux. Economically, this is much more interesting than commoditization, but it won’t have much of an impact on Microsoft’s business which is largely driven by corporate purchases of new servers, workstations and desktop productivity applications. Users of these business applications aren’t about to migrate to ultra-portables anytime soon. Alas, I also do not see these folks migrating to Linux anytime soon. The reasons there are many and varied – a good topic for a future post.
Instead of the demise of Microsoft, I see further gains by Linux here at the expense of special-purpose proprietary operating systems. As the manufacturers of consumer devices realize the tremendous advantages in cost and time-to-market that come with using Linux over cooking up their own O/S, we’ll see Linux (and other complimentary open source software packages) in more consumer devices in the coming years. The implications here are many and all are good for the consumer; lower prices, cross-device compatibility, and easier mods among them.
See, my opinion lies somewhere in between-
The feedback I’ve been hearing about Vista constantly is that it’s ‘bloated’. Hell, the Mac/PC ads were even making a joke of Microsoft’s new OS, and how much it was under-whelming consumers. This annoys me to no end- MS holds most of the market share, but they make products that get that sort of feedback just after launch? Apple, at least, throws a lot of effort into their products, but is only now learning that a lower price point will get people to buy their stuff more.
What has this proved? That the average consumer likes simple, low-priced computer products.
The Asus Eee is proving that people are capable of learning to use something other than Windows, out of the box. It’s small, it’s cheap, and it works- and the word of mouth advertising that can come from that will hopefully make an impact.
Do I think it will ultimately kill Windows? Not in my generation! But the point is that hopefully Asus and other companies can take enough of Microsoft’s market share to make them a little more savvy about creating an operating system that doesn’t make people groan with displeasure and go out and buy a Mac.
Many of my friends are doing that, and I’m seriously considering it….
Nathen,
Great points. I really believe that it is not the OS that drove the Asus EEE pc purchases, but that age old consumer electronics mantra… Price Point, Price Point, Price Point. Manufactures learned a great lesson from the Asus and the OLPC XO. People will buy a low cost laptop, the price point is ~$300. Now, does the OS matter? Not really if Asus can ( and I bet they will ) get a deal from Microsoft to supply the OS for that same price point, it will sell.
The good news is that because of that price point more people were exposed to Linux, and some will stick with it, because it is a viable alternative.
The hardware is the commodity, the software is not, and won’t be until all OS’s inter-operate. That will drive prices down. Much like all gasoline from different stations run your car the same way, when OS drive your computer the same way, then the OS is a commodity.
I thought commoditization was the process whereby product become undifferentiated. I think its fair to say that operating systems are, in some ways,becoming less different from one another. Here goes econ 101:
As Operating systems becomes more mature, the features that differentiate the operatiting systems will become less important. This is commoditization, when the products become interchangeable. The price will become zero because the marginal cost of software is zero. The cost is unrelated to commoditization. These are two separate issues. For example commodities can be expensive…like oil, or differentiated products can be cheap….coke and pepsi are cheap. Keep trying though you will get it someday.
Marginal cost of software is ZERO
You noted — “If the operating system market was commoditized, then by definition cost would be the primary point of competition. As we can see, more than a few consumers and businesses chose costly, proprietary operating systems from Microsoft and Apple over free Linux alternatives. I don’t believe that a stronger argument need be made against the thought that the operating system is commoditized.”
Nathan, you forgot a core design criteria methodology; Application to be run –> OS required to run them –> OS purchased. People don’t buy pipes for pipe, they buy pipes to have water. The same can be said for OS’s. People buy the OS so that they can run applications like MS Office and Quicken on it.
Don’t be fooled by the numbers you quote. Back out 5% for inflation. Then back out a 17% premium for the cost of Vista over the previous OS, XP. If anything Vista is flat. What you really want to quote is unit sales not revenue. Nor does $$ sales growth have a correlation with commodization unless you know what is the unit sales cost for an OEM license.
Thanks for the comment John - very good points indeed. I really like your pipe analogy; this is one of the issues I was alluding to when I stated that corporate users would not be moving to Linux any time soon.
My point in quoting the revenue figures from Microsoft & Apple was to illustrate that this is still a highly differentiated market. So much so, that the primary players are making billions even while competing against a product that is being given away.
“80% of which came from their OEM channel”
well, quite!
thats exactly the problem. with the eeepc and the likes, consumers are for the first time having the option of purchasing a computer with a non-MS operating system, or at least not having to pay MS for an operating system (even if they choose to run a different one thereafter).
The OEM-lockin monopoly is an _absolute scandal_, but once this is over, MS will fade away deservedly!
moving from having a choice and competing against a billion-dollar marketing budget with something free and good will take some time!
When something is a commodity it’s common, used by many people, available to everyone, and priced so that most people can get it if they really want it. Often commodities are important for every day survival.
In a sense, Heathkit, Zenith, Commodore and Tandy commoditized the operating system.
Webster says:
1: commodify; specifically : to render (a good or service) widely available and interchangeable with one provided by another company
2: to affect (as a brand or a market) by commoditizing goods or services
The above named companies commoditized the OS by the second definition. Linux commoditizes the OS by the first *and* second definitions.
Examples:
Bike tires
the ford model T
digital audio workstations
x86 hardware (once IBM opened the spec)
Technically, the “Operating System” was commoditized with the Commodore64, Tandy, Atari, Texas Instruments and Zenith computers and has remained so since the late 70’s. Once these came out, anyone with $200 could have a working computer, including the operating system. While very minimal and a lot like a BIOS these still meet the definition of what an OS is.
DOS commoditized the OS on x86 hardware and really started the ball rolling.
FOSS operating systems have taken this commoditization further, not created it. Linux is available to anyone regardless of financial means, like air and you can compile and run a given piece of linux compatible software on any company’s distro, which meets the first definition. That being said, Linux and BSD *did* commoditize “UNIX like Operating Systems”, just not “Operating Systems”.
Couldn’t resist this because I was there. I watched computers and OS’s become commoditized as it happened. Commodity does not mean free, just common, cheap, and sometimes interoperable.
By contrast z/OS running on a mainframe is not commoditized, though often large companies use mainframes to run commodity OS’s as virtual machines.
If OS’s were not a commodity, only the very wealthy could afford them.
-Viz
Thanks for the comment Viz - very thought provoking.
While the pioneering companies you mentioned certainly did make computing accessible to the masses, that’s not the same as commoditizing the market. Operating systems are widely available, but they are not “interchangeable” - and this is critical for commoditization. You seem to be discounting this aspect of the definition. Ask yourself - if Windows, OS X, and Linux were truly interchangeable, why would anyone buy OS X or Windows when they could get Linux for free?
Your last statement implies that in order for a good to be affordable, it must be a commodity. Further, your comment in general, suggest that any good which is affordable is a commodity. This isn’t the case at all. Gasoline, strawberries, electricity, and natural gas are all commodities. Yes they are affordable, but this is a function of the cost to produce and deliver them to market. Other commodities are quite expensive: silver, gold, and diamonds for instance. In a free market, low margins are a necessary effect of commoditization, low prices are not.
Beer, soup, breakfast cereal, and soda are also all very affordable, but they are not commodities because consumers do not see the alternatives in the market as being interchangeable. Think back to the last time you bought a beer or a soda. Did you buy a generic brand? Most folks don’t - even though by doing so they could save 40% or more on the purchase. The distinguishing characteristic of a commoditized market is that the companies playing there compete primarily on price. That doesn’t mean the price is low, it just means that the margins are small.
Finally, while I disagree with you that the desktop market in general is commoditized, my gut instinct is to agree with you that UNIX-like operating systems are now commodities. Some research there might be in order though.
Story added…
This story has been submitted to fsdaily.com! If you think this story should be read by the free software community, come vote it up and discuss it here:
http://www.fsdaily.com/Community/Is_Linux_Commoditizing_the_Desktop_OS_Market…
Operating systems and a lot of other software are very, very complex products made of tens of millions of lines of code. I doubt they will ever be interchangeable commodities in the sense milk or natural gas is. The interactions between different pieces of software and hardware are just that much more complicated compared to those material commodities.
Where there are more or less standardized interfaces and software truly can be interchangeable, Linux and other free systems have largely taken over. Proprietary POSIX systems have been forced either to the margin (AIX, UnixWare) or to change to a free license (Solaris). Obviously the big more-or-less standard interface driving commoditization of personal systems is the web. I’m writing this using the latest beta of Firefox 3, and I’d say Windows and Mac OS X don’t have anything in the browser department that Linux does not have. The number of systems and users that only need a browser is growing, and where cost matters, free systems will take over these users.
The big interfaces that Microsoft owns are the Office file formats and the Exchange server. If these areas are to be ‘commoditized’, I think it’ll be done by large corporate and institutional users switching to open standards, not individual home users. OpenOffice seems to have a growing list of large institutional adopters, but I’m not sure it is significant compared to Microsoft’s installed base.
An excellent point Mike. What you are saying about the network being the commoditizing factor in the market jives well with Jonathan Schwartz’s blog Commodities, Railroads and How Sun Monetizes Java (http://blogs.sun.com/jonathan/date/20040712).
[…] What price would the market demand? […]