Firefox Site Visits Up 237%

prostoalex writes "Nielsen//NetRatings, a top Web reporting and metrics agency, started tracking the Firefox Web site in June 2004 and has announced 237% growth since then. Nielsen tracks Firefox Web site visits, not downloads or usage patterns, but it notes that "Men accounted for 71% or nearly 1.9 mln site visitors, compared to the women who comprised 29% or the minority population who visited in March 2005.""

Oddly enough...

[kwoo]

The most popular browser/OS combination to my sites (which are Unix-oriented) is Firefox/WinXP.

Firefox/Linux is actually in second place. IE of various flavours on Win32 is third.

Certainly not what I expected to see before starting the sites, that's for sure -- but it's roughly the same mix on each one.

Re:Validity of the article linked to?

[Xshare]

This article is about visits to mozilla's website, not people using mozilla browser.

Re:Nielsen?

[wdd1040]

Neilson has connections to a few very high traffic sites and they either A)allow them to view the statistics on their logs, or B)allow them to put a 1x1 gif on their website for usage stats.

Re:Nielsen?

[Anonymous Coward]

Except you simply cannot get accurate statistics on browser usage that way. Oh sure, lots of people try, and lots of people publish their results, but it's all built on a foundation of sand. The web just doesn't work that way.

I can't seem to find a description of Neilsen's methodology, but if they are reporting male:female ratios, then they are using data beyond things like web bugs, and so I would be inclined to trust these figures far more than other organisations, including any data we could cull from our own log files.

Re:Downscale

[digidave]

Each person does not need to go to the site more than once, then just use the browser's built-in update mechanism to update to new versions.

Looking at it like that, it means that most of these visitors are brand new to the site rather than returning visitors, thus meaning that they have increased their reach several times more than 300%.

Nielson/Netratings has Java/Javascript code that runs on their customers' web sites to report traffic back to them (RedSheriff). If Firefox put that on their site they would be able to tell just how many of these visitors were returning from previous months.

Re:Opera is more efficient.

[Anonymous Coward]

Most important to me, once a page is loaded, accessing it is instant in Opera. Say you click on several links. You can go "back" any number of pages and each one instantly appears, without reloading or any of that inconvenient stuff.

That's because Opera follows the HTTP 1.1 specification, and other browsers, including Firefox, are non-compliant. RFC 2616 says:

History mechanisms and caches are different. In particular history mechanisms SHOULD NOT try to show a semantically transparent view of the current state of a resource. Rather, a history mechanism is meant to show exactly what the user saw at the time when the resource was retrieved.

Opera and Konqueror get this right. Firefox doesn't, and isn't standards compliant, because it checks for updates in many circumstances.

Re:Oddly enough...

[kai.chan]

From the statistics of my personal website, I am getting 30% non-IE hits consistently. Because my site have a wide range of materials that doesn't really cater to a specific crowd, I have drawn the conclusion that although a large portion of businesses still use IE, Firefox usage percentages looks to be over 20% for home users, which is a significant gain over a year ago when it was at ~5% for my site. Does anyone have any statistics that separates the percentage of corporate machines running Firefox versus home machines running Firefox?

Re:Calling Home

[2*2*53*4127]

And how does NetRatings know the gender of the visitors?

i wondered that myself. probably an opt-in deal, like the neilson TV families who allow their viewing habits to be tracked and mapped against their demographic?

Re:Nielsen?

[ScentCone]

You can visit the Nielsen-NetRatings [nielsen-netratings.com] site and bone up a bit, and if you want, grab a a PDF of their corporate brochure [nielsen-netratings.com] which mentions that their techniques include the usual image tag bugs, but also techniques just like they use when they do TV ratings: interviews with "recall" information, journals, and other (for us web folks) seemingly unlikely approaches. It's all about doing sanity checking against traditional (and easily polluted) web stats. Big companies like to have their facts audited and tested by alternate methods, and Nielsen's been doing it for a long time with other hard-to-measure stuff.

Re:Oddly enough...

[Anonymous Coward]

For my small personal Website, IE6 is currently at 63%, down from 75+% a year ago, and Gecko browsers account for 21% (16% of those are Firefox) of hits.

It's nice to see a mixture of browsers (Konqueror, Safari, Opera, even the occasional text browser) in my logs too.

Re:Nielsen?

[digidave]

Um, no. Neilson bought RedSheriff last year. RedSheriff is a web analytics and data collection service that many sites pay for. The site would drop a piece of code onto their pages, including some Javascript, Java applet and a 1x1 gif.

From there the site owners would have access to an online reporting tool that is quite good.

AFAIK, RedSheriff didn't share or use their customers' site traffic logs for any purpose other than to report back to the site whose logs they were. Nielson may have re-jigged their privacy policy to allow it.

About the gender statistics...

[Neopoleon]

It looks like there's a lot of confusion about the gender data gathered, mostly along the lines of "How'd they do that?"

I know it sounds crazy, but I went ahead and visited the the Nielsen site and read up on their strategy. I realize this goes against the techie tradition of never RTFM, but that's a risk I was willing to take.

Turns out they use a "holistic" approach to their data gathering. Everything from "server side blabbity-blah blah blah" to conducting surveys, hiring people to browse, and tracking ad clicks.

I'm guessing that the gender comes from the surveys, but I don't want to upset anybody who might be really excited about a new gender-aware version of HTTP.

If you want to read up on this stuff yourself, you can check out some info here:

http://www.nielsennetratings.com/mktg.jsp?sectio n= ps

Click on a few products to see the range of apps/services offered. You'll see where all this data comes from.

Re:Opera is more efficient.

[Anonymous Coward]

And RFC 2119 says:

4. SHOULD NOT This phrase, or the phrase "NOT RECOMMENDED" mean that
there may exist valid reasons in particular circumstances when the
particular behavior is acceptable or even useful, but the full
implications should be understood and the case carefully weighed
before implementing any behavior described with this label.

To summarize: not following a "SHOULD NOT" in any RFC does not make something noncompliant.

Re:Oddly enough...

[kwoo]

The most popular browser/OS combination to my sites is FireFox/WinXP too - but I'm the only one that uses them, so that information doesn't mean anything :)

For most of that time, I've used Mozilla on Solaris/x86 to access the sites.

Parent would be worthy of its Informative mod if there was some scope to its claim (I could mod as overrated but I'd rather actually find out what sort of number of people we're talking about here, because its pretty impressive if Firefox is the #1 browser on a decent-sized site!)

My original comment is overrated -- had I known it would be rated so highly, I would have put it in some context. Daily hits fluctuate between 50-500, and content page views between 40-300. Not a big site by any means. Monthly unique hosts is in the order of 500. About 40 TLDs are represented in each month's logs, and about 10% of the unique hosts can't be resolved back to a domain.

Hope that helps.

Re:This is to be expected...

[ozmanjusri]

Thing is, Firefox defaults to the Firefox website!

Not only that, but the default page on the Firefox site has a Google search field right in the middle. Most of the people I know (including IE users) have set Google as their start page. With Firefox, there's no reason to change. Smart.

Re:Just be happy

[gnarlin]

Sorry to burst your bubble, but Firefox is not covered by the GPL. It's under the MPL (Mozilla Public License).

More stats, but probably realistic

[mjtg]

As luck would have it, I went through my organization's web logs just the other day for info on browsers. Here's a summary of what I found, for anyone who's interested.

12 months ago, IE accounted for a steady 94% of hits. Gecko-based browsers (Netscape 6+, Mozilla, Firefox) accounted for 3%. Netscape 4 had around 1.5% of the hits, Safari just under 1%, Opera about 0.5%, and Konqueror 0.1%.

Firefox started registering in my logs around July, when the Gecko share jumped to 4.3%, rising steadily to 5.7% in October. In December Gecko jumped up to 7%, and is currently around 8.2% (March-April). Firefox now represents about 80% of Gecko-based browsers. The number of non-Firefox Gecko hits (ie. Netscape 6+, etc) has remained more-or-less steady.

IE's decline matches Firefox's rise - by October, it was down to 92%. IE now rates around 87% of hits on our site.

Safari has increased to about 2.5%. Netscape 4 has (finally) declined to virtual insignificance. Sadly, Konqueror has also declined steadily, maybe 0.03% in a good month (looks like a lot of Konqueror users have switched to Firefox too).

These stats come from an Australian state government website that receives about 3 million hits per month. The site is not technology-oriented, and about half of the hits come from overseas, so I believe that this is a reasonably good sample of browser use.

Re:Downscale

[asa]

More people download Firefox from the www.mozilla.org homepage than from the slightly less visible www.mozilla.org/products/firefox page that Nielson was apparently measuring.

- A

Re:This is to be expected...

[asa]

"Thing is, Firefox defaults to the Firefox website!"

Have you used Firefox? It defautls to http://google.com/firefox.

- A

Re:Of course, Firefox is the default home page...

[Compumyst]

Just to let you know, the default start page [google.com] is located on the google website, and would not be counted towards a hit on the firefox website.

Re:This is to be expected...

[Mars Ultor]

Actually I have used it, thanks for asking. Since it was called Phoenix.

I do realizet that the google page is now the default for firefox. but at least a year ago, it was still pointing here:
http://www.mozilla.org/products/firefox/cen tral.ht ml

In fact, I just installed Gentoo 2005.0 and the default for Firefox is still the mozilla.org site.