[joneeboi]

While generally being a lazy dog when I should have been studying, I decided to look up head-fi's traffic graph.

What does this graph mean? Why do you suppose there's that huge lift in '06 before a huge decline? I'm guessing that was the first national meet, but why did everyone ditch after that first national meet? Was there that much tension and anticipation for the meet starting at the end of 2004? Had meet attendees missed the announcement of the 2007 meet? What is that little dip about at the right end of the graph? What do you suppose this next one indicates?

Details at head-fi.org - Traffic Details from Alexa

Do comment.

 

[NeObliviscaris]

Shouldn't there be like a 2 week dip for the downtime? Or does the pageholder still count as it being a visit?

 

[Wmcmanus]

My answer to all of those questions is: I don't know. Not that they're not interesting questions, but the graph seems rather odd to me. I would have thought it would have graphed pretty much like a straight line.

It seems to me like that is (more or less exactly) what has happened over the years in terms of the key parameters (i.e., number of new members, number of posts, number of threads). Thus, I would have expected this same kind of straight (upward) path to have occurred in page clicks (which is what I assume this data represents).

Of course there may be a week here or there that doesn't follow the general pattern, but this graph is showing mega trends that seem to fly in the face of some other significant data that we've observed. The long and steep decline from early '06 to early '07, in particular, doesn't make any sense to me. I wonder if they had some data capturing problems, or in some significant ways redefined their measurement techniques, etc.

The only explanation that comes to mind for this is that about that time (early '06) was when Jahn was told he could no longer post on Head-Fi all day long while at work. I think that might explain it as well as anything. Where we've seen upward spikes since then are probably highly correlated with his vacation time.

 

[TheMarchingMule]

Quote:

Originally Posted by Wmcmanus /img/forum/go_quote.gif The only explanation that comes to mind for this is that about that time (early '06) was when Jahn was told he could no longer post on Head-Fi all day long while at work. I think that might explain it as well as anything. Where we've seen upward spikes since then are probably highly correlated with his vacation time.

Priceless!

Where is that picture of his posting team when you need it?

 

[jude]

Alexa rankings are based on views by people with an Alexa Toolbar. Up until recently, Alexa traffic was pretty much only counted on Internet Explorer browsers, as the Alexa Toolbar is only available for Internet Explorer. Recently, however, an offering for Firefox became available (it's not a full toolbar, but a status bar add-on). Still, most Alexa traffic is being counted from Internet Explorer visits. Why is this relevant? Whereas Internet Explorer is still the dominant browser, in terms of browser share worldwide, it is actually a rather distant second-place for Head-Fi.org visitors, and has been for a while, with Firefox trending up (even now), and Internet Explorer trending down. For example, since the site went back up on November 25, 2007, Firefox has a Head-Fi visit browser share of 52.63%, versus Internet Explorer at 32.34%. I believe that, in terms of worldwide browser share, Firefox is more along the lines of 10% to 15%, and either flat-trending or even slightly dipping as of late; with Internet Explorer at 80'ish percent. The Alexa graph with respect to Head-Fi.org would probably be at least somewhat attributable to that.

Nevertheless, Alexa still ranks Head-Fi as the most popular in its categories (at least the categories I'm aware of that it's in):

1. HiFi
2. Audio
3. HiFi Community
4. Audio and Video

Prior to the outage on November 10, 2007, we were at our highest traffic levels ever, with just under 300,000 unique visitors per month, and record-high page views. After the outage, we lost some traffic, with our traffic levels right now about where we were in early spring 2007, but gaining.

In other words, I can't fully explain the Alexa graph shape up there, but our numbers (versus the typical website) may be being underrepresented due to our disproportionately high Firefox browser share here.

The 2006 spike doesn't surprise me, however, as it was at that time (the time of the First National Head-Fi Meet in New York in 2006) that we set our all-time record for concurrent visitors (I can't remember what it was now, as it was reset after the outage--it's probably cached somewhere, though).

 

[OverlordXenu]

Isn't Alexa generally considered useless?

Publictivity.com Blog » Alexa is The Most Useless and Misleading Web App of All Time
Alexa Says YouTube Is Now Bigger Than Google. Alexa Is Useless

Edit: Beat by Jude.

 

[jude]

Quote:

Originally Posted by OverlordXenu /img/forum/go_quote.gif Isn't Alexa generally considered useless?...

I still pay some mind to it, because others do, too. For my purposes, though, the only Head-Fi.org traffic data I really pay solid attention to is the data I get from Google Analytics. I like tag-based analytics (versus log-based), and Google's purchase of Urchin worked out nicely, and provides very complete site metrics for the price (free).

 

[fordgtlover]

Here's a graph showing the outage:

It would seem that the initial graphs are from the reach panel. I took this one from the page views panel.

 

[Jahn]

Santa might have been joking but I did sorta join Head-fi right when things started escalating a few years ago, and at that time i used internet explorer. and i was indeed given the lockdown practically on the day of the first National Meet (i remember chatting with Jude about that, he said "it wasn't us! it was your workplace!") and soon after that I switched to firefox at home.

IT. ALL. FITS.

Alexa, How Could You??!?

 

[ken36]

Good stuff.

 

[mbriant]

http://www.head-fi.org/forums/3122063-post83.html

 

[cerbie]

Quote:

Originally Posted by mbriant /img/forum/go_quote.gif http://www.head-fi.org/forums/3122063-post83.html

The same trend exists, Jude has given an explanation that can fit for any of these forums, and most related ones.

I can compare head-fi, SPCR, AT forums, HardOCP forums, diyaudio.com, and CPF, and...

While the exact data differs, the trends are the same for them all. Head-fi appears worse on the page view count, but the trend is still there for the others. CPF, FI, has a couple oddities in 2007, but the forum was also down for awhile, and split up. Pre-07, the trend appears the same.

Alexa's data gathering being questionable looks to me like a really good explanation.

Notice all have a major jump in reach towards early 06, with Head-fi being the highest of the smaller ones, a dip going into Q4/06, with a bump as 07 comes in, a dip towards summer, then a slight-but-obvious rise (excepting CPF, as explained).

Usage data worth trusting for actual use is that coming from people with access to server logs.

 

[3DCadman]

The IE vs. Firefox explanation is the most likely, but the Jahn explanation is the most satisfying/humorous and that's the one I'm accepting

 

[Edwood]

Alexa is about as useful and accurate as Neilsen ratings.

-Ed