W3Schools Records February Mozilla Usage Share of 9.0 Percent
Friday March 5th, 2004
jrs_66 noted that global Mozilla usage is growing fast, according to browser statistics from W3Schools. The latest figures state that Mozilla had a 9.0 percent usage share in February, up from 8.2 percent in January. The W3Schools data indicates that Mozilla's usage share has more than doubled in the last twelve months.
Is this the suite browser, or a cumulative count of all Gecko-based web clients?
Well it can't be all geckos because Netscape 7 is listed separately. But I would guess it is all geckos that are not Netscape 6/7. But IE is slowly losing market share. IE 5 lost 1.3% but IE 6 only gained 0.8% from January to February. So IE lost 0.5% but Mozilla gained 0.8%. Moz didn't get it from Opera, and only got 0.1% from old Netscapes so it must have gained the other 0.2% from the unlisted other/unknown category. And/or maybe they added more geckos like kmeleon, galeon, etc. to be under Mozilla.
But all in all a consistent gain even if it's small is good news. Because the switch won't happen over night. So the slow spread through word of mouth and outside exposure thanks to articles and reviews is working. This trend also could be following the trend of increasing use of linux.
On my low traffic domain IE is only at 78.5%, Moz 7.7%, Unknown 6.9 %, Netscape 4.5 %, Safari 1.2 %, firebird 0.5 %, Opera 0.3 % for february. I'm not sure where firefox is being reported. My guess is Unknown. But I was surprised to see firebird listed on it's own. My host put in a new stats package in December. IE was 84.7% in January. Firebird was 2.1% in January so it must have gone down due to the name change.
Another interesting thing is I'm getting is a small number of IE 7.01 hits. What's the deal here? Spoofed UAs?
#4 Re: Need help? Do it yourself
by lacostej <email@example.com>
Saturday March 6th, 2004 12:57 AM
IE7: could it be that? <http://webperso.easyconne…82-ExtendInternetExplorer> Read on Daniel's web site a couple of days ago.
#5 Re: Re: Need help? Do it yourself
Saturday March 6th, 2004 8:38 AM
I doubt it. That code is something you put on your web page to force IE to work properly. It's not something you put in your web browser.
#10 Re: Re: Need help? Do it yourself
Monday March 8th, 2004 2:13 AM
No, if you look at the server log, the complete UA is : Mozilla/4.0 (compatible; MSIE 7.01; Windows 98) And the referrer is always : <http://www.thebestofnet.c…/topsecret.cgi?6743452643>
It's just hidden advertising.
#12 Re: Re: Need help? Do it yourself
Monday March 8th, 2004 2:39 PM
Microsoft mentioned that it won't be making any standalone IE and that it would be embedded in Windows from now on. So, what's your looking at is something like Window Explorer or Window Paint for example. No one is going to really know about the IE version once it get embedded into Window, it would be nothing but of a memory.
>Another interesting thing is I'm getting is a small number of IE 7.01 hits
I read before that some early alpha versions of Windows Longhorn had leaked out. Perhaps some people are trying them out, since Longhorn is supposed to have a new version of IE.
...that's what I am seeing on my site (it's caters to developers though) as well. This month was interesting. FireFox started out strong in its first month and squeaked by IE 5.x (9.39% vs 9.27%). Mozilla's share dropped a little bit from 6.85 to 6.69, but its drop in terms of share of Gecko browsers has dropped dramatically from 60% to about 30%. FireBird's (that's FB 7.0) share dropped from 2.44 to 2.19. Looks like people are slowly abandoning both Mozilla and FireBird. Netscapel 6,7 is hovering at about 0.7%. I guess, the only people using it are those who need it to retrieve email from netscape.net or those who have to use it from work. Other browsers that use Gecko engine, like KMeleon, Epiphany, etc... are nearly invisible. For what it's worth, whatever traffic I get from Linux (less than 2% of total), it's nearly all Gecko.
So all in all, Gecko share is around 18% vs 10% last month. I would guess that some of the jump can be attributed to people trying out FireFox. The reason I say this is because for the 1st 5 days of March, FF's share is back on par with Mozilla. Will have to wait and see whether people stick with FireFox.
And IE 6 was the king of the hill with 65%.
I support open standards as much as anyone on mozillazine but how realistic are these statistics. I don't have anything to prove my point of view, but I feel that 9% is way too optimitic, furthermore how much of those 9% are concentrated on non-mainstream sites? Can we really say that some mainstream site like say yahoo.com would get 9% gecko hits?
Isn't this based on visitors to W3Schools? Probably more Gecko users there than otherwise, since they are all into web development etc. Wouldn't the Google Zeitgeist be better for these statistics?
#11 Please stop reporting worthless figures...
Monday March 8th, 2004 3:04 AM
Yes indeed, it's entirely pointless to look at statistics from one particular relatively small site, especially one targeted toward web developers.
I would very strongly encourage MozillaZine staff not to report statistics from any individual small site (where 'small' means 'not comparable in size to Yahoo, Google, MSN, etc.'), however favourable they appear to Mozilla. IMO such information is sent in either by misguided Mozilla supporters or, worst-case, by the sites themselves trying for a little publicity.
I also agree that the best public statistics are those from Google Zeitgeist. The most recent figures are for January 04. In terms of pixels on graph and then percentage share (of that total), those figures are: IE6 100 (68%), IE5.0 15 (10%), IE5.5 12 (8%), Other 9 (6%), Gecko 7 (5%), IE4 3 (2%).
So IE overall has about an 88% share, with Gecko at 5%. Gecko has actually been trending *very very* slowly upward on their graph (taking about 5 months to rise 1 pixel), while IE6 has been rising quickly (on the back of declines in IE5.x, obviously); in the most recent update it appear as if IE6's upward trend has slowed, but frankly there's not enough data to tell.
I wish they'd give their figures directly so that I didn't have to measure the graph :) By the way, I didn't try to 'second-guess' anti-aliasing, I just took the highest coloured pixel in each case.
#13 Re: Please stop reporting worthless figures...
Monday March 8th, 2004 2:43 PM
So, how would you tell if the statistic from Google, Yahoo, etc. is what it reflect? Anyone can use Google like 100 times today and 0 the next day. So, the statistic there is not much of a difference as it look.
"So IE lost 0.5% but Mozilla gained 0.8%. "
--> which is the small representative unit? I mean, it is like some political survey: you ask the political choice to 3000 people but you get an uncertainty of 3 or 4 % ... so a gain of 0.8% on one month have to be set in relation with the incertainty percentage of the statistics ... how many unique visitor had visiting w3school???