It seems to have passed without huge fanfare, but on April 20th 2006, the Alexa rankings for tech sites around the globe suddenly doubled, literally overnight.

The Day The World Got Geeky

Ian Clarke writes, "Did the world get geekier around April 20th?"

Some kind of change to how Alexa calculates traffic – or did something happen to so dramatically affect these four websites? Perhaps a bunch of geeks suddenly decided to become Alexa users.

Alex Walker noticed the phenomenon as early as the 27th, titling it "The Great Internet Spike of 2006."  Alex noticed that only tech sites seemed to benefit from the Alexa bump:

sites with no notable ‘tech-skew’ (i.e. CNN.com, EBAY.com, etc) have either held firm or been shuffled backwards by the sites bubbling up around them.

Have Alexa changed their statistical algorithms?
Is this a temporary anomoly?
Which sites lost out the worst?

Even Digg users have been speculating:

hayseed: also possible that a crapload of people installed the alexa toolbar at that time as a result of some sort of promotion ...

hottuna: Alexa is so incredible unreliably its strange that the site hasn't been shut down.

jarcoal: exactly. there have been several articles reporting major traffic boosts in the last few months. i'll bet this is all some stupid glitch.

When Elephants Fight

I've been looking at this a bit, and I'm convinced it's no Alexa glitch.  Rather, the Alexa population did, in fact, change overnight.

Here's the funny part: Digg users were the scalawags who did it!

Before I tell you why this happened, let's take a look at a graph:



Here you see slashdot (red line) moving along toward mid April, trending slightly downwards.  Digg (blue line) appears to have a consistent trend as well, though their user base is steadily growing.  Then, around mid April a singularity occurs as both lines meet.  BANG!

This isn't a spike, folks.  It's a massive, towering wall of geekdom.  The same wall hit Wired, BoingBoing, Engadget, Gizmodo and just about every other tech-bent site on the planet.

Now, I know what you're thinking...  When those two lines merged, all the Digg users and the Slashdot users put aside their differences, started mingling, got friendly, and had litters and litters of geeky babies, thereby doubling the geek population in a single evening.  I was thinking this too, but then my wife reminded me about our own baby.  I can attest, from personal experience, that even for geeky humans the typical gestation period is about 38 weeks.

Now, I'm not totally ready to rule this theory out, but I figure we have 'till about late December or early January 07 before we see the SlashDigg baby spike on Alexa.  Even then, it's only going to be the real hackers that push their way out and immediately trade in their umbilical cord for an 802.11 hotspot.

Lost my train of thought -- back to what happened in April.

The truth all falls back on what Ian and hayseed had speculated: suddenly, one fine April day, a _lot_ of techy folks downloaded and installed the Alexa toolbar.  So many, in fact, that their statistical influence shuffled the rankings down of huge sites with no nerd bias.  The interesting thing is that it had nothing to do with an Alexa promotion, a gang of rogue Digg hackers, or any intentional funny-business.

In a way, though, it actually was because of the convergance of those two little lines.


The Masses Converge

In mid April, Digg approached the worldwide reach of Slashdot, formerly regarded as the most widely read tech site on the net.  People had been speculating about the event for some time.  Then, on April 19th, "it finally happened," writes ejoyner.  "Digg overtakes Slashdot" is posted and the story instantly hits the Digg front page.

Hundreds of thousands of Digg readers follow a link to Alexa en-mass to see internet history being made.  Some percentage of them install the Alexa toolbar that very day, and the head-count that this percentage represents is huge enough to significantly skew the Alexa demographic from that point forward.


Observations

This seems to be a pretty unique event, and I'm left with a couple of thoughts.

First, if you are the devious webmaster of a really large site, you can significantly skew the Alexa rankings in your site's favor simply by driving your users to Alexa with an interesting story.  Ironically, if Slashdot had announced the story that Digg had surpassed it in ranking, we would likely have seen the Slashdot ranking spike to the top.  By not heavily publicising their own bad news, they effectively allowed Digg to claim a substantially higher Alexa ranking!

What if no announcement had been made?  What if no Digg users downloaded the toolbar and Alexa was left to continue charting internet popularity with their former sample base?  Did Digg really overtake Slashdot in the end?

Ultimately, nobody but Alexa could really know for sure, but if we were to assume for a moment that the Digg and Slashdot graphs continued their pre-20th trends and approached eachother at the end of April, we may have seen something like this (adjustment translated below actual data):



Just look at the correlation between those two graphs over the mid April to mid June period!  For as much as they complain about eachother, it would appear that Slashdot and Digg users represent the exact same population.

Also, in this hypothetical scenario, Digg really only began reaching more readers than Slashdot in late June.  A smart Slashdot editor could probably capitalize on this and declare right now that Digg has officially bested them, with a prominent link to Alexa to prove it.  Did I just say that out loud? ;)


Geek Rank Your Site

One final interesting take-home from this whole discussion is that you can use the April 20th "geek day" to gauge the geek factor of your favorite sites.  How much spike did you see on geek day?  Did you see a drop?  Make sure to tell us about it in the comments.

Trackback URL for this post: