There's lots of Google Chrome OS buzz around the Web today, much like what we saw 9 months ago when Google Chrome browser was announced. The headlines are, predictably sensational.
I have a question though. When all the hype has blown over, how will we know if Google Chrome OS is a success? What's the measure? What kind of penetration do we expect in the next few years?
With the Google Chrome browser there was an awful lot of speculation that it would destroy Firefox (or IE) in no time at all. We're more than 9 months out from that hype and the Chrome Browser has destroyed nether Firefox nor IE and has only just beaten out Opera in terms of usage share with about 2% of the global browsing market.
So what is success in Google's new "war" with Microsoft? I'd like to offer a measure of success that's at least somewhat rooted in reality.
If Google Chrome OS has 2% of OS usage three years from now, I think it will be safe to say it's been "wildly successful."
What follows is some back of the napkin math, some educated guesses, and some wild speculation that led me to that 2% figure.
Here's my thinking. Google has stated that they're targeting netbooks first. 2008 saw about 10 million netbooks sold. Considering the ongoing global recession, 2009 will probably see that double to 20 million sold. Let's imagine that it grows to 30 million in 2010, 40 million in 2011 and 50 million in 2012. This is wild speculation, and I suspect it's on the high side, but I'm trying to imagine the best possible landscape for Google Chrome OS.
So, from the time Google gets into this market, a year from now, through their first two years of Google Chrome OS distribution, there would be approximately 85 million netbooks sold. What percentage of that do I think Google will represent? Well, that's a difficult question to answer but I think a generous guess would be that they could take 1/3rd, or just over 28 million in sales away from Microsoft. That's kind of an ugly number, though, so I'm going to round it up to an even 30 million.
The installed base of Web-connected computers is about 1.2 billion and if growth continues at about the same pace, it will be about 1.5 billion by the summer of 2012. (See this post for more on that number.)
And we do the math. 30 million divided by 1.5 billion gives us 2%.
If Google does everything right with their Chrome OS and they manage to acquire a full third of netbook sales in their first two years, then we'd see Chrome OS represent about 2% of Web-connected computers by the summer of 2012.
If you accept that, then you sort of have to admit that the headlines around the Web today are pretty over the top.
This goes to a larger point, I think. New markets on the Web can emerge an grow really quickly. There's lots of opportunity for something like Facebook to take over in just a few years. But that's not really the case for PCs and desktop software. The installed base is just really, really large, and the growth and upgrade cycle are much much slower than with Web services.
Firefox has been the most successful piece of desktop software to ever challenge Microsoft's offering. We started the effort 10 years ago and finally arrived at a successful product 5 years ago and in the 5 years since we shipped that product, we've managed to gain about 300 million users and a quarter of Web browsing usage.
Apple has been the most successful operating system to challenge Microsoft ever and they've managed in the 8+ years of OS X availability to grab only about 5% of the global OS installed base.
It's just not fast or easy to move a market that's more than a billion large. Anyone that thinks that major change can happen in months, or even a couple of years, doesn't understand this space very well.
Posted by: Mike Beltzner | July 8, 2009 9:45 AM
I guess so. But after 2 years Google might have the product that can compete with Windows in large scale, and the story will be different. Google might just iron out some disadvantages of its system in that time, and still keep its advantages. So from that perspective it makes sense that there might be something big behind this. So it has potential, and whether it will be something big or nothing special is now up to Google.
And while your metrics are true, I believe that it is biased by browser statistics. So in free markets you are more likely to follow install base as metrics, but in economy it is current market share (market share of units sold last month/year) that counts. And this will be much better for Google than installed base share.
But I believe that there is dark side of this story - Microsoft was not too opened, but Google won't give others the chance to touch anything. I can switch to Linux or Mac and keep most of my data there. But I can't from gmail to hotmail. And my freedom is limited just in chosing web application (and somehow it is likely that Google's application will be better integrated than others, so choice might be an illusion :) ).
I really belive that it is right time for Mozilla to take much of leadearship in some OS effort. Willl it be Moblin or something else, I have no idea. But in the near time, we might just not be able to install browser on any consumer platform.
Posted by: Ivan Ičin | July 8, 2009 10:39 AM
2% is small in installed based, but huge in revenue loss for MS.
In WWII the Germans were defeated far from home, Stalingrad and El Alamein.
Lucas
Posted by: Lucas | July 8, 2009 11:22 AM
Not so sure the tech-savvy and privacy aware people (which is a tiny minority) will go for this at all.
Chrome OS: Google can mine my email, my location, my searches and now my entire computer? No thanks.
Posted by: David | July 8, 2009 11:28 AM
"I can switch to Linux or Mac and keep most of my data there. But I can't from gmail to hotmail."
Not true. Actually gmail is one of the few popular mail accounts which gives POP access to its mails from very early on. I read all my mails from a mail client and not from browser. If you want to move to other webmail accounts, use their POP access feature (most of them have) and get all mail from gmail to that account.
Jay.
Posted by: Jay | July 8, 2009 12:09 PM
Awesome post.
I'm wondering. Will Google allow for the installation and use of web browsers other than Chrome? :)
Posted by: Ken Saunders | July 8, 2009 12:59 PM
Jay, you miss the point - the point is to use old messages on new web application. And it is impossible now. Even if you use desktop application as a bridge between those webmails (and there even won't be such application on Chrome OS as far as I understand). And who would switch and lose all his old mail?
Posted by: Ivan Ičin | July 8, 2009 1:37 PM
I doubt Google Chrome OS can achieve even 2% of the OS market, since that'd be twice of the market share that the entire Linux family currently has. And Linux has been here for more than one and a half decade already.
And I surely don't like the "everything in the cloud" thing, it will mean Google will have control of all your data, and your whole life. It looks to be a step towards those big evil mega corporations in those scifi movies, like Weyland-Yutani.
Posted by: soshimi | July 8, 2009 6:10 PM
Ivan, the fact that Google provides IMAP access with Gmail means you can - in fact - transfer your mail wherever you want. I know because I did it in the reverse when I migrated my ~10 years worth of mail into Gmail. If I want to take it all out and go to some other service that provides IMAP access I can. If I want to export my contact list to a CSV I can.
Say what you want about privacy or data mining, but Google has always been pretty good about data portability and preventing lock-in. It's tough to fault them on that front.
Posted by: Jeff | July 8, 2009 9:04 PM
(this is getting off-topic, but I'm curious so I'll allow it.)
Jeff, can you please give me the simple steps I need to replace Gmail with Yahoo mail. Are there two or three clicks that will magically move all my mail and contacts (and various preferences) from one service to the other?
On the desktop, this proved to be really easy. I downloaded a new client, launched it, OK'd the "migrate" option, and I was up and running running. it was literally one click to migrate everything over.
Can you tell me how I do that one click (or two, or even three clicks) to migrate everything from Gmail to Yahoo Mail or Live Mail?
- A
Posted by: Asa Dotzler | July 8, 2009 10:23 PM
and the same question for moving from Google Apps to Zoho. I installed Open Office on one of my machines and, like magic, I could access and work with pretty much all of my Microsoft Office file. It just worked. How do I effect the same kind of change if I want to start using Zoho to manage all of the documents I was previously managing with Google Apps.
- A
Posted by: Asa Dotzler | July 8, 2009 10:26 PM
Gmail does have a one-button magic service to import contacts and old mails from Yahoo Mail, Hotmail and etc. It's under the settings -> Accounts and Import tab.
You can also import contacts and old mails from gmail to Yahoo mail with Yahoo mail's Options -> Import Contacts service.
For Google Docs and Zoho, you can export the documents into local files, and then import those files to those online services. It's quite a hassle, sure, but I guess in the future they can work just like those web mail import/export services if they cooperate more.
Posted by: soshimi | July 8, 2009 11:55 PM
You might want to mention that Google currently provides the Mozilla foundation with 85% of their revenue). If Google ever cut Mozilla loose the game could certainly change.
Personally I don't like the idea of using a browser provided by an advertising company and the ongoing Google Updater issue *is* proof that they're evil!
Posted by: Casual Observer | July 9, 2009 3:00 AM
Asa, as others have said, it's not *easy* to migrate but it's possible, and I think that's all we can really expect on Google's end. I don't know what options Yahoo offers for import, but if they offer IMAP I'd tackle the problem by setting up Thunderbird with both the source and destination IMAP accounts, and then manually drag & drop everything from one account to the other. There are also various local backup solutions out there for Gmail based around automating the same process.
(Sorry for dragging the comments off topic, but it bugs me when people say something I've done in the past is "impossible".)
Great analysis of the ChromeOS situation, by the way. Seeing ChromeOS's best case scenario in terms of worldwide marketshare really puts it all in perspective.
Posted by: Jeff | July 9, 2009 5:07 AM
Like I have said, at least for Gmail and Yahoo, it's actually not difficult, but as easy as clicking a link and then enter the other account login and password, and all your old mails and contacts will be migrated.
For Chrome OS, I still think Asa is overestimating it. To put it all in perspective, that means in two years Google will be able to accomplish twice what the whole FOSS community in the world could barely accomplish in one and a half decade!
I don't think Chrome OS will be able to magically triple Linux shares in two years.
Posted by: soshimi | July 9, 2009 5:27 AM
shoshimi, I'm not making a pure prediction here. I'm saying what I think would have to happen for this Google effort to be considered successful. The picture I'm trying to paint is that "IF Google did EVERYTHING right and EVERYTHING worked in their favor..."
There are plenty of different outcomes possible, most of them I think are Google doing less well than "wildly successful." It's possible, for example, that netbook sales don't increase over the coming years. Heck, netbooks could mostly disappear about the time Google's ready to release this thing. Who knows.
I just wanted to point out that even if Google kicks ass with this effort and does awesome, all of the "Microsoft is dead talk" is just silly.
- A
Posted by: Asa Dotzler | July 9, 2009 7:15 AM
Good analysis, Asa. You're absolutely right that anyone saying 'Microsoft is dead' needs a swift boot to the head.
That said, even with only 2% of the market you can effect a lot of change. As Blizzard likes to point out all the time, minority market-share software can influence majority-share software makers to pick up their game. Chrome the browser is keeping Moz on its toes even if not actually picking up much market share, Linux forced Microsoft to up their commitment to stability and quality, and OSX appears to have forced them to focus on usability (if early reports about the improvements in Windows 7 are accurate.) So Google comes out ahead even if ChromeOS gets merely 1%, as long as that 1% forces Microsoft to take netbooks seriously, and makes MS work harder to improve performance and web ease-of-use- those things will mean more web users and more ads served. That, in my book, counts as pretty smart strategy on their end.
Posted by: Luis | July 9, 2009 8:49 PM
This one is funny:
http://www.theregister.co.uk/2009/07/08/orange_and_ie6/
You have to fine people, for not using Firefox.
Posted by: Lucas | July 10, 2009 5:52 AM
Did you get your Mac OS figure from the median count in Wikipedia's usage share article?
http://en.wikipedia.org/wiki/Usage_share_of_desktop_operating_systems#Summary_table
I always thought the methodology the Wikipedians used to get that number is very questionable... calculating the median of multiple sources when these sources aren't equivalent at all in audience and sample size (ex. bundling W3C numbers with OneStat, among others) doesn't produce a representative result.
I bring this up since Net Applications have Mac OS usage share creeping close to the 10 percent mark for quite a few months now. As you mentioned somewhere in your blog before I also think their statistics is more indicative than the other usage share trackers on the 'net.
Posted by: Paul | July 10, 2009 11:46 PM
Just to note that Chrome OS is far from original. Actually, Microsoft was first to publish the paper on this:
http://www.readwriteweb.com/archives/introducing_microsofts_gazelle.php
Posted by: Ivan Ičin | July 12, 2009 12:47 PM
"I bring this up since Net Applications have Mac OS usage share creeping close to the 10 percent mark for quite a few months now. As you mentioned somewhere in your blog before I also think their statistics is more indicative than the other usage share trackers on the 'net."
Yes Net Applications is better than the other usage share trackers, but still it is still next to useless.
From its ISP tracking stats, we know that it takes its data mainly from North America and Europe, mostly ignoring the rest of the world. For example, in real world, Asia has over 30% of the world's internet users, while North America has less than 20% of the world's internet users. But from Net Applications ISP tracking, Asia has less than 1% of the total ISP tracked, while North America has over 40%
That means Net Applications' statistics does not represent the real state of the internet in the real world at all.
Considering the extremely low penetration rate of Mac OS in a lot of Asian countries, I'd guess the real marketshare of Mac OS is much lower than the 9.8% shown by Net Applications. I'd bet my money on somewhere around 3-5%, but then there's currently no way to tell for sure, since all those available online stats don't take account of the Asian market even remotely realistically.
Posted by: Hikerseven | July 12, 2009 11:47 PM
I grew up in China and had been to a few other Asian countries. I can tell you from my personal experience, you might not choose to believe me cuz my personal experience tells me Open Source supporters are nuts (I like Open Source, I just can't stand the cultists who like to post "10 reasons XXX is a MS killer" before anybody in the world actually knows what that thing actually looks like , let alone get to play with it, and despite the fact that they had posted the same damn thing 1 zillion times before, each time they were proven wrong. Geez, get some grip). Anyway, in Asia, on desktop, nobody except a few very dedicated professionals use linux, even among computer science majors, windows absolutely dominates. Chinese government has been trying really hard to push home-grown linux distro and indigenous office suite (which is free for personal use, by the way) via governmental procurement policies, but I can tell you that the second those machines are actually placed in the offices, the first thing the workers who are supposed to use these linux boxes is to reformat the hard drive, if they know how to do it, they'll do it themselves, if they don't, they'll find somebody else to do it for them, and install windows and MS office. Everybody inside Chinese tech industry knows that these Chinese OS/office suite companies are nothing more than leeches surviving almost solely on (the office suite company isn't cuz they are pretty big in on-line game, which reaps big bucks for them) governmental support. It's not that their product are so lousy that they didn't stand a chance, they didn't stand a chance because the dominance of MS windows and MS office is almost complete.
Posted by: Jack Wang | July 13, 2009 2:22 PM
Great post!
Posted by: Ephilei | July 15, 2009 5:14 PM
sadsadsa sadhhd yuejnd dusahd fdjd ifmdsjm, i8djs 2+6 45 jstgd ufjdujf isjd ajds
Posted by: theson26 | July 24, 2009 6:11 AM
Great analysis, Asa. Really adds a ton of context and perspective.
I'm pretty happy about the announcement - the more movement towards "web-like OS" the better, as I agree with Google's statements that for application developers, the web should be an available platform akin to OSX or w32 or Linux. There's a lot of work to be done in this space, and having a formidable player like Google with their resources will be helpful, especially if they are fully open with their work. I'd love, for instance, to know if they're bypassing X, replacing X, etc.
It's also good to know that there's pressure on OS vendors to not ignore the emerging market of computers that aren't built for heavy client side OSes, but rather slim ones that speed people to the web.
So: yes, it'll take time, but I think it's a step in the right direction and can't wait to see how the web community at large is involved.