February 04, 2005
Tipping Firefox across the chasm.
There has been a lot of interest in the issue of browser stats of late, with Firefox's meteoric rise to prominence (putting paid to my prediction early last year that simply building a better browser was not enough to build a sizable market share.)
So because I had nothing better to do on a rainy Sunday morning (well, I had lots better to do but...) I've done a quick analysis of the overall browser market share of visitors to westciv.com
For a very small company (there are in essence two of us here) Westciv generates reasonable traffic. This year we are likely to exceed 1.5 million unique visits, and serve up perhaps as many as 10 million pages. Note that the visitors to our site are largely reasonably experienced web developers. I'll discuss the significance of this shortly.
These are the browser stats for everything over .5% of traffic. The stats don't add up to 100% because we have not included any browser with less than .5% of market share, and I didn't bother factoring out robots, and unidentifiable browsers.
Nonetheless, the stats are, as far as I am aware, reasonably accurate, and take into account browser spoofing and the like. And the sample size is big enough, hundreds of thousands of visits in each of the three periods below, as to make the them reasonably significant.
So below are the market share by browser brand, with percentage change on the previous period where available.
Browser Stats 13 May 2004 - 11 July 04
- MSIE 54.20%
- Mozilla 18.79%
- Safari 6.11%
- Opera 3.13%
- Netscape Navigator 2.15%
- Lynx 0.55%
Browser Stats 12 July 2004 - 30 October 04
- MSIE 49.82% change - 8%
- Mozilla 21.63% change +15%
- Safari 5.87% change -3.9%
- Opera 2.83% change -9.6
- Netscape Navigator 2.27% change +5.6%
- Lynx 0.72% change + 30%
Browser Stats 1 November 2004 - 18 Jan 05
- MSIE 42.87% change -13.95%
- Mozilla 25.63% change +18.49%
- Safari 6.71% change +14.3%
- Opera 2.56% change -9.54%
- Netscape Navigator 1.97% change -13%
- Lynx 0.4% change -44.4%
- MSIE change -20.9%
- Mozilla change +36.4%
- Safari change +9.8%%
- Opera change -18.21%
- Netscape Navigator change -8.3%
- Lynx change -27.27%
Note that MSIE represents all versions of IE, but particularly IE 5 and IE 6.
Mozilla includes Firefox, Mozilla and Netscape 7, in essence anything that identifies itself as using the gecko rendering engine (with our focus on CSS and web development at westciv, above all my interest is in rendering engines, not browsers per se).
Interestingly, about 34% of our current users use IE6, and 9% use IE5. If you go back to May of last year, the split was then 44% and 9.5%. So, the IE market share is actually falling much more significantly than the raw totals would suggest, particularly among early adopters. My guess is that IE5 users are "rusted on" either because they are obligated by corporate policy to use IE5, or because being at the end of a dialup connection they aren't inclined to upgrade to IE 6. IE 6 users show much less "loyalty", which if I were Microsoft, unless I had a very good reason not to, I'd be very worried about.
Now, as mentioned above, we aren't talking about a broad cross section of web users here. Westciv's content is of little interest except to web developers, who I'd guess are among the earliest adopters of new web browsers.
But my guess is that these stats reflect the day to day browsing habits of web developers, it's the browsers they use when looking for information, not just something installed for testing. So what we are seeing is that among web developers, a very sizable percentage have adopted a Mozilla based browser for their day to day browsing in the last 9 months or so.
Which got me to thinking about where these trends are heading. Will Firefox supplant IE as the dominant browser, just as IE replaced the once dominant Netscape?
A couple of reasonably well known books, that I've found have over the years given me some interesting tools for thinking about problems like this came to mind as I thought about the issue - Crossing the Chasm, by Geoffrey Moore, and the Tipping Point by Malcolm Gladwell.
Let's start with Moore's "chasm" analysis of how new technologies get adopted.
In a nutshell, Moore observes that the adoption of a new technology is "quantum" like. There is no steady adoption of a new technology. Rather, two kinds of users, "enthusiasts" and then "visionaries" adopt a technology in its infancy, but the real challenge for a technology is for it to "cross the chasm" and be adopted by pragmatists, by the "mainstream".
Moore makes the point that It's not uncommon for a new technology to be adopted by a great many of these early adopters, but fail to "cross the chasm".
I don't think Firefox/Mozilla (I'll use the term Firefox from now on) is across the chasm yet. I'd say web developers (particularly those interested in "new" stuff like CSS) are a mixture of visionaries and pragmatists.
And I'd say Firefox is in the "tornado", Moore's term for the rapid takeoff of a new technology when visionaries start adopting it in significant numbers, but before it has crossed over to more mainstream adoption.
Note that being in the Tornado does not guarantee that a technology or product will "cross the chasm" (or, as Malcolm Gladwell would say, will "tip"). (For more on this, see Gladwell's justly famous piece for the New Yorker on Cool Hunting, or just read the Tipping Point). As Gladwell describes it, the tipping point is
the name given to that moment in an epidemic when a virus reaches critical mass. It's the boiling point. It's the moment on the graph when the line starts to shoot straight upwards. AIDS tipped in 1982, when it went from a rare disease affecting a few gay men to a worldwide epidemic. Crime in New York City tipped in the mid 1990's, when the murder rate suddenly plummeted
Will Firefox Tip
So, will Firefox tip? How would we know if it had? Has it tipped already?
My gut feeling is that it has not hapened yet. The stats from W3Schools, westciv, and other developer oriented sites definitely suggest a major adoption by developers, the web's "early adopters" and "visionaries", not just in terms of testing and playing with Firefox, but in using it in their everyday use.
The difference between westciv's stats and those of w3schools I'd suggest reflects the different kinds of developer who use these sites. W3Schools users are more likely to be less advanced developers than westciv's (this is not a criticism of anyone, simply an observation about the different kinds of content the sites provide). I'd predict Slashdot's stats would show an even more marked adoption of Firefox over IE (which in part would reflect the larger user share of Linux among slashdot readers). Taco, any chance of publishing your stats?
But the rate of adoption in more mainstream use of Firefox appears considerably slower. This is not necessarily something to be discouraged about (if you, like me, and most web developers, want to see at least a significant non-IE browser at play to push innovation, and stop stagnation in browser development).
The Chasm and tipping point models of adoption would suggest that wild growth among the early adopters is a necessary precursor to more mainstream adoption. But it's not sufficient.
What will it take?
So what's is it going to take for Firefox to go mainstream?
The Wikipedia Chasm entry summarizes Moore's analysis of the challenges confronting a new technology when attempting to move into mainstream adoption like this.
Moore suggest[s] techniques to successfully cross the chasm, including choosing a target market, understanding the whole product concept, positioning the product, building a marketing strategy, choosing the most appropriate distribution channel and pricing.
With Firefox, some of these seem to be non issues. To be widely adopted, browsers must be free, and Firefox is. That takes care of pricing.
The distribution channel is an interesting issue. You download browsers, right? So the distribution channel is "the internet". However, its clear that Internet Explorer's rapid rise to market dominance in the late 90's had not a little to do with it's being on the desktop of every new windows system installed.
Now, it seems unlikely that Microsoft will be doing this favor for Firefox any time soon (nor Apple, given their investment in Safari, for that matter).
So is Firefox stuck with "the internet" as its distribution channel?
One of the most important ways in which users upgrade their browser is when they upgrade their internet connection, or change to a new ISP.
With significant growth in Cable and Wireless internet connection predicted in the near term, ISPs mindful of the support costs of Internet Explorer - associated with its vulnerability to spyware and other malware - may choose to make Firefox their browser of choice.
AOL's recent announcement of a Mozilla based Netscape 7 upgrade is essentially an example of this.
And I wonder just how much malware and spyware related system problems are costing Dell, HP, and other Windows system vendors in terms of tech support, in an industry where margins are razor thin? Even just answering the phone, and telling people it's Microsoft's fault could be costing these vendors a lot of money. Putting Firefox on the desktop as the default browser might just save them a small fortune.
What about a target market for Firefox? What's that about anyway? Surely the Mozilla Foundation want everyone to use it right? So the target market is "everyone who uses the web".
The idea behind choosing a target market is to recognize the difficulty of a single product being all things to all people. Choosing specific target markets allows developers to focus on the needs of a particular set of users, and marketing to articulate the benefits for a specific kind of user.
Reading their product pages, and considering their recent New York Times advertising spot, it appears that the Mozilla Foundation has chosen a target market of "everyone" for Firefox.
Is this a good strategy? (as opposed to goal, I think that for a free browser, unless you are widely used, you are likely to quickly become irrelevant. So the goal of any free browser is to be as widely used as possible, but it is often not a great idea strategically to try and get everyone to adopt your technology right away.)
Now while Firefox does not appear to have chosen a focussed, target market, interestingly, a couple of focussed target markets have apparently chosen or are in the process of choosing Firefox.
The first are people who have some kind of commitment to Free and or Open Source Software. These are your definitive hard core early adopters. Their motivation is at times anti proprietary software, or more specifically, anti Microsoft software. Or it might be a more idealistic motivation, that software should be free (either as in beer, or as in speech). The classic example is the slashdot reader.
The second group is web developers. Web development sites, like westciv, as we have seen, see use rates of Firefox at 10-20 times higher than more mainstream sites tend to report. Which can't be a bad thing.
As I've argued earlier, this does not mean Firefox has tipped. But, it does mean Firefox has a chance of "tipping". To take Gladwell's analogy from the tipping point a little further, let's imagine the adoption of Firefox as a kind of epidemic, like the flu. In order to catch the flu, you need to be exposed to someone who has it, and is contagious. A disease that is only infectious for a very short period of time, even if everyone who comes into contact with it will catch it, will not necessarily become an epidemic, because not enough people may be exposed to it for the epidemic to "tip".
Compare the flu - which is mildly infectious, but once caught, its carrier remains infectious for at least several days, with say Ebola, which is highly infectious, but remains infectious for only a very short period of time (usually because it kills its victims quickly) and the warning signs of which are so distinctive that potential victims can identify the risk and take precautions to prevent infection, unlike with cold and flu viruses, where they are infectious even before physical symptoms are obvious. Gladwell calls this second vector of epidemics "stickiness".
Now what on earth has this got to do with whether or not Firefox is ever going to really take off?
By analogy, to take off, Firefox needs to be both "infectious" (people have to be prepared to give it a go) and "sticky" (that is, it must remain infectious for a reasonable period of time like a flu, not a very short time, like Ebola).
Is Firefox Infectious?
Firefox certainly appears to be infectious. We've seen that it is spreading like wildfire among certain populations - based on our statistics, among web developers at about 50% per annum, which would give it a 50% plus market share in little over a year. Anecdotally, I know of many "mainstream" technology users - those who don't keep up to date with developments at all, who simply "use" their computers and the internet - who also find Firefox and even the full Mozilla application a welcome improvement to Internet Explorer when exposed to it, so much so that they stay with Mozilla/Firefox.
But those infections I know of in the mainstream have been because of the evangelism of early adopters - who might be fed up with fixing the malware related woes of their friends ad family each time they visit, for instance. The Mozilla Foundations challenge will be to infect more mainstream users.
But is Firefox just a Fad?
At times people speculate that Firefox might be a fad. When I was at high school, an epidemic of "two up" (a particularly Australian form of gambling) swept the school play ground. At lunch literally 95% of the kids played two up in small groups. It was incredibly infectious. I have never seen anything like it since. But none of those kids were playing a couple of weeks later. The fad was not sticky enough. That and the priests banned it because it was gambling.
I don't think firefox is a fad like this. Unlike "Pushcasting" (a virulent short lived fad of the mid 90s that almost cost News Ltd. 400 Million dollars) stats and anecdotal evidence suggest it is sticky enough that users once they start using it, continue indefinitely.
Firefox looks pretty "sticky" too. It seems that among those populations who use Firefox, like the web developers who visit our site, that this is not a fad. The continued solid growth of Firefox, almost entirely at the expense of Internet Explorer 6, indicates that users are not flirting with this browser, but using it significantly as part of their day to day use of the web.
If we return to Moore's Chasm analysis, to get Firefox to cross the chasm to mainstream adoption, he would suggest that Mozilla Foundation "build a marketing strategy" and understand "the whole product concept".
What can Gladwell's thinking on epidemics teach us about these two challenges?
When Moore suggests "build a marketing strategy", what goal should it have? Based on Gladwell's analogy, it should be to "make Mozilla infectious" .
Meanwhile, understanding "the whole product concept" means understanding how to make Mozilla "sticky".
Get people to use Firefox. And once they start using it, keep them using it, wanting to use it. Needing to use it.
Making firefox Infectious
Internet Explorer's increasing tiredness (not having been updated significantly for several years now, with little if any improvement promised for the indefinite future), the growing general understanding of the importance of security, and IE's vulnerability to spyware and other malware, gives increasing impetus to web users to find alternatives. While your average user may simply grin and bear these increasing frustrations, their early adopter friends and family are much more likely to take matters into their own hands and cut the gordian knot by simply installing Firefox in place of Internet Explorer.
ISPs, as well as Dell, HP and other vendors who shoulder increasing support costs from users calling with systems slowdowns, and virus, spyware and malware related problems might start thinking about saving themselves a pretty penny by making Firefox the default browser on their Windows systems.
So, should the Mozilla foundation concentrate on this negative aspect alone in marketing Firefox? It certainly appears to be a significant focus of their branding of Firefox (on the firefox homepage the first quote - from USA today - and two of the first product bullet points mention security and popup blocking) but not the only one.
But Firefox also urges us to "rediscover the web", a better aspiration than "get rid of the your current browser because of all these problems".
If I had any advice for the Mozilla Foundation when it came to marketing Firefox it would be that they really separate out the two very distinct groups of user they currently focus on - visionary early adopters, and the mainstream - rather than lumping them together when promoting the application.
Visionaries largely want complicated, complicating (and wonderful) stuff like plugins, toolbars and the like. The appeal to mainstream users of Firefox is almost the opposite - simplicity, security, ease of use, and maybe even a little fun. A bowser that "just works". Taking back the web. There you go Mozilla, you can have that one.
Right now, Mozilla, who do you really want to use Firefox? Where is your energy going? You've got plenty of traction with developers and early adopters. Perhaps you should stick with them for just a little bit longer, and keep working on usability issues for more mainstream users.
Making Firefox sticky
The Mozilla browser feels more like a whole heap of cool technologies than a single integrated application. Classic early adopter stuff. With Firefox though, the Mozilla Foundation has concentrated on the whole product approach. By integrating emerging technologies like RSS, and making Firefox a more seamless way of working with leading web technologies like Google, Firefox is not just a collection of really good pieces of technology, but rather, a complete product in itself.
Now Mozilla work harder on making Firefox even stickier.
You've integrated RSS in the browser. Now refine an RSS interface that "just works". Be the ones who made RSS mainstream. Own this emerging aspect of the web, like you own the idea among web developers of standards compliance (perhaps the key reason for Firefox's support among developers).
Improve bookmarking, make it smarter and more intuitive, so that you aren't just playing catchup with IE, but everyone else starts playing catchup with you.
Integrate even more seamlessly with the major web services that make the web easier to use, like you currently do with Google - with the Internet Movie Database, EBay - you know the stuff that those mainstream users use the web for.
It's clear that Firefox is in the Tornado. But not all new technologies cross the chasm to mainstream adoption. The danger at present for the Mozilla Foundation, and it's well wishers to see the rapid growth of Firefox use as implying Firefox's necessary mainstream success. It doesn't.
But many of those in the web development world, like me, wish it well. Internet Explorer 6 was released in 2001, coming up to half a decade ago. The web still needs innovation. And a successful Firefox might not only deliver that, but spur others - Apple, Microsoft, Opera - to do so too.
February 4, 2005 | Permalink
TrackBack URL for this entry:
Listed below are links to weblogs that reference Tipping Firefox across the chasm.:
» Browser stat from JD on MX
Browser stat: Here's another snapshot, of another type of site, this one catering to "reasonably experienced web developers"... recently they're seeing 43% IE, 26% Mozilla among visitors. There's also good material here on how audiences differ, how thi... [Read More]
Tracked on Feb 5, 2005 11:51:45 AM
» "Tipping Firefox Across the Chasm" from Ravensky.org
A long and informative read, but if you have time for it, I’d suggest reading it. Very interesting. Haven’t read all of it myself (yet), but it seems to be talking about the world’s new adaption to browsers other than Internet Explore... [Read More]
Tracked on Feb 5, 2005 4:21:25 PM
» Tipping Firefox across the chasm from .redSPLASH - Blog
John Allsopp takes a detailed look (Tipping Firefox across the chasm) at the current state and the possible future of the web browser ecosphere. [Read More]
Tracked on Feb 5, 2005 10:28:16 PM
» Does Firefox == Mainstream? from weissblog:
Much has been said over the last few months about the Mozilla Firefox browser and how it seems to be gaining market share. Some people speculate that Firefox will soon capture the market similar to the way Internet Explorer crept... [Read More]
Tracked on Feb 6, 2005 2:39:34 PM
» Taking Firefox Past the Tipping Point from 8 Ways to Sunday
Westciv looks at the trends behind the Firefox hype, and concludes that the upstart browser still needs help to gain mainstream adoption...... [Read More]
Tracked on Feb 7, 2005 7:51:40 AM
» Has Firefox "Tipped"? from Firefox
dog or higher: Tipping Firefox across the chasm. This bit below is part of a very good analysis of Firefox market share and market penetration. I highly recommend clicking through to the entire article if you're interested in Firefox market... [Read More]
Tracked on Feb 25, 2005 9:56:23 AM
Tracked on Oct 16, 2005 5:39:08 PM
Amazing stuff, thanks.
Blake Ross certainly does understand the "making Firefox stickier" point -- take a look at his ¨Firefox religion¨ post (https://blakeross.com/index.php?p=9).
Improved bookmarking and some neat history management would be nice, I agree with you on that.
Posted by: Rimantas | Feb 5, 2005 12:04:09 AM
So I should come to visit you a bit more often with Lynx , eh ?....
Joking aside, interesting analysis. I agree with you, Firefox remains the best opportunity we have to break open the eco-system, and have a bit more diversity. We're not yet there though, although in most cases when I helped people discover alternatives to IE windoze, they're quite happy with it. But we're not yet there, and there is a bit too much a mentallity of copying/catching up with IE win.
Better personal data managment (bookmarks and history, but also stuff like passwords and the like) is one of the things that could help.
And better technology to install/manage Firefox in a coorporate environment. As far as i know, big steps have been taken there, with a network installer available now.
Posted by: Philippe | Feb 5, 2005 1:47:22 AM
I think there are two reasons why IE because mainstream.
You mentioned the first - for home users, it was pre-installed on the desktop and/or distributed by ISPs. This was important because home users don't install software unless they have to.
The second reason for IE's popularity is it's adoption by businesses. IT departments are not content with pre-installed software, they willing to install what is required. IE was and is a good development platform for Intranet applications. Corporate developers could easily write plug-ins (in VB for example) or use a host of pre-built plug-ins like the ones that come with Outlook. Developers could also integrate the IE component into their own applications written in VB, C++, etc. Overall, IE gave more development choices to corporate developers than Netscape or early version of Mozilla.
Mozilla Foundation has been focusing solely on home users. In my opinion, in order to become mainstream, the Foundation needs to focus on the hearts and minds of corporate software developers.
Posted by: Vlad Alexander | Feb 5, 2005 4:27:50 AM
As previously mentioned, it's an unfortunate fact that many average users will probably never be changing from their default browser. This means that IE users will account for a significant portion of the traffic at most sites we develop (excepting perhaps our personal/special interest sites) for the forseeable future, and we will always have to give IEs flaws some consideration. Thankfully, IE will need less special attention as people buy new computers with new OSes already packaged with IE6, however their will still be considerations to make for that browser.
From all that I've read, Microsoft has stopped developing Internet Explorer (and why not, I can't imagine it's a big money-maker, and it certainly isn't the dealbreaker that makes people choose windows over another OS.) I think the best case scenario would be that they abandon their rendering engine, and replace it with gecko for a final version to be released with their next OS. Perhaps the developers could just send Microsoft the occasional updates for inclusion in the automatic update system. Imagine the bliss of nearly every person browsing with the same engine. Lots of people would lose money for time spent on IE-proofing, but that's the only downside as far as I can see.
That level of cooperation is wishful thinking, I'm sure, but just imagine.
Posted by: Scott | Feb 5, 2005 8:16:30 AM
Nice insights; I'm one of those people installing Firefox on my friends' broken PC's. Note that MS Win Update still requires IE. Log on with Firefox and you get scolded.
Posted by: Steve Torrente | Feb 6, 2005 10:40:33 AM
I am trying to be a carrier, being, and finally getting my family to be Mac based, I can only agree wholeheartedly with the 'make it mainstream and simple to use' philosophy. I have got my 76 year old dad blogging and using a mac but I think trying to explain why he should use Firefox rather than Safari is a step too far. Mind you if he was still on windows...
Posted by: Julian | Feb 7, 2005 5:49:10 AM
Echoing Julian - I'd consider myself an early adopter, I've been using Firefox for a couple years now - but now I've gotten my parents using Firefox instead of IE. They love it, they don't really know why, they know it works better and keeps their computer working better (or, not working badly). The tip will come when this starts spreading across different kinds of groups with different demographics - the generation gap is one chasm that Firefox seems to be leaping now.
Posted by: Kelly | Feb 7, 2005 10:33:20 AM
We run a large golf website and our stats show that 91% of our audience still use some sort of IE browser and Firefox sits just below 5% despite us pushing it as the browser to use.
The IE (main variants) split over the past four months is:
Microsoft Internet Explorer 6.0 - 82.93%
Microsoft Internet Explorer 5.5 - 3.99%
Microsoft Internet Explorer 5.0 - 3.71%
So it seems that perhaps the non-techno types have not caught on yet and golfers are certainly in that bracket...well and truly. It is not a good sign for us as we want to move to a fully CSS site but do not want to spend our days hacking for IE6.
Posted by: John | Feb 7, 2005 2:02:26 PM
What about courting the websites that are taking the brunt of the spyware assault? Banks. They're the ones that have to deal with the hassles of customers getting their accounts drained or identities stolen because of stolen passwords. And perhaps suprisingly, banks seem to pop up frequently among the sites that only try to support IE.
It's probably difficult get in to such a conservative organization, but surely there are some clueful bank sysadmins with clout who realize how negligent Microsoft has been, and how much *they're* having to swallow all the risk.
Posted by: Eric | Feb 8, 2005 4:46:37 AM
When do Firefox extensions come into play? This has a major impact on support/distribution. I doubt many of the major ISPs or manufacturers would switch to Firefox without the ability to lockout the installation of extensions, or at the very least be able to access the extension list to determine what a user has installed, otherwise supporting the myriad of possible functionality is almost impossible. But is THAT more costly than malware issues with IE???
As for "Be the ones who made RSS mainstream." I'm not sure that is such a key aim? But then I've yet to see a similar article to this on the benefits of RSS and how it needs to be 'tipped' (presuming that need exists).
And finally, the bookmark issue. What price a move to a delicious model?
Great article, gave me plenty to think about, thanks.
Posted by: Gordon | Feb 8, 2005 8:15:18 PM
Here are stats from various websites that are not affected by web developers. Just stats collected from multiple sites with counters on them.
Mozilla accounts for only 1% of all browsers.
Posted by: Ottawa | Feb 17, 2005 8:33:19 AM
while I don't argue at all (quie the opposite) that regular users have anything like the adoption of Firefox that developers do, those numbers look all wrong in lots of ways. They get a smaller percentage of users using IE5 than we do! No mention whatsoever of Safari. My guess is that its a very superficial check on the browser at best.
Posted by: John Allsopp | Feb 17, 2005 9:05:08 AM